Psychiatrist Jie Zhang developed the Continual Activation Theory of Dreams in 2004. This theory states that during sleep, information is transferred into long-term memory. Dreams are a by-product of this data transfer process. The Continual Activation theory says that dreams are caused by random memories that the brain retrieves in order to keep all parts of working memory continually active during sleep.
Zhang’s theory combines aspects of Hobson and McCarley’s Activation Synthesis theory with aspects of Mark Solms’ work. Like Hobson and McCarley, Zhang says that dreams are stories that the brain makes up in order to make sense of random pieces of information. Like Solms, Zhang says that dreams can happen during both rapid eye movement (REM) sleep and non-rapid eye movement (NREM) sleep, and that REM sleep itself does not cause people to have dreams.
Declarative and Procedural Memory
The Continual Activation theory distinguishes between declarative memory and procedural memory. These two types of memory are processed in different areas of the brain.
Declarative Memory
Declarative memory is the memory of facts – things that you can declare, or discuss in words. It can include facts you learn from books, facts that other people tell you, as well as facts that you learn from your sensory experiences – “Jane was wearing a red dress on Tuesday.” We use our declarative memory consciously – we try to memorize facts and we make conscious efforts to recall facts. Declarative memory is also known as explicit memory.
Procedural memory
Procedural memory is the memory of procedures – how to do things. Procedural memory is not conscious and is often difficult to put into words. It involves things that our bodies do without our thinking about how we do them. Procedural memory can include things such as knowing how to tie your shoes, how to play a musical instrument or how to drive a car. Procedural memory is also called implicit memory.
Sensory, Working, Temporary and Long-Term Memory
Zhang created a model of how we process memories. This model says that memories pass through the following four memory stores: sensory, working, temporary and long-term memory. Our brains filter our memories as they pass from one memory store to the next. Only a small portion of the data that is stored in sensory memory gets passed all the way to long-term memory.
Each of these stores is further divided between a store for declarative memory and a store for procedural memory, each of which is located into a different area of the brain. This means that memories are stored in eight different areas of the brain.
Sensory Memory
Sensory memory is the buffer zone for every piece of information we receive from our senses. Data is held in sensory memory for only milliseconds up to a few seconds. Once we receive information from our senses, our brains almost immediately filter out information that is unnecessary.
The information that is stored in sensory memory is similar to the information that is stored in a photograph (although sensory memory includes sound, smell, taste and touch, as well as sight). A photograph includes every minute detail of a scene, without regard to what is important or unimportant. Sensory memory works the same way.
Working Memory
Information is passed from sensory memory to working memory through the process of attention. Working memory is the area of memory where we manipulate information. For example, suppose you are trying to solve a math problem in your head, and you have to break the problem down into smaller steps. You store the answer to each step in your working memory. When you see someone and then try to associate a name with their face, you are using your working memory.
According to Zhang, when your working memory isn’t working properly, you fall into a coma. Therefore, working memory is probably located in the brain stem, because if you have lesions on your brain stem you will fall into a coma, but you won’t fall into a coma if you have lesions of your frontal lobe or if you’ve had a lobotomy.
Temporary Memory
Temporary memory is a limited capacity storage space for memories. Information moves from working memory into temporary memory, but only stays there for a while, and is then either deleted or moved to long-term memory. When your temporary memory is overloaded, it is extremely hard for you to learn anything new.
Long-Term Memory
Long-term memory is the place where memories are stored permanently. Long-term memories can last up to a lifetime.
How Memories Are Processed When You are Awake
When you are awake, you receive information from your senses which is transferred to your sensory memory. Then, using attention, you move some of this information into your working memory. To process this information in your working memory, you must retrieve related information from your temporary memory and your long-term memory:
• “I see a person. My temporary memory tells me that I was just introduced her. What is her name?”• “I’ve just been given a math problem which involves adding two three-digit numbers whose last digits add up to a number greater than 9. Can I find the instructions for how to do this in my long-term memory?”
Once you process the new information in your working memory, it is sent your temporary memory, where it is saved.
When you are awake, you save all of your new memories in temporary memory. You do not save any memories into long-term memory, which is in retrieval-only mode.
How Memories Are Processed When You Are Asleep
Information is transferred from your temporary memory to your long-term memory when you are asleep. Declarative memories move to your long-term memory during NREM sleep. Procedural memories move to your long-term memory during REM sleep.
When you are awake, you can’t clear out your temporary memory. This is because your temporary memory is busy helping your working memory process the large amounts of data that you receive from your senses. Your temporary memory needs some time when it is completely shut off from the environment, so that it can clear itself out and make room for new information. When you are asleep, your temporary memory is shut off from the outside world, so that the necessary housekeeping can take place.
When you are asleep, your senses work more slowly and your sensory arousal thresholds are increased. Your temporary memory is in retrieval-only mode. You can’t record any information, such as dreams, into temporary memory.
How Memories Are Processed During NREM Sleep
During NREM sleep, your brain moves declarative memories from your temporary memory to your long-term memory. Your declarative working memory retrieves data from your declarative temporary memory, compares it with related files from your declarative long-term memory, and then deletes any unnecessary or duplicate data. The updated information is then transferred to your declarative long-term memory.
During this time, you may have night terrors when your brain is processing extremely frightening memories. Night terrors are known to take place during NREM sleep.
If someone wakes you up during NREM sleep, you may remember what Zhang calls a Type I dream, – a thought-like dream that is not very vivid. A Type I dream that takes place during NREM sleep consists of declarative memories that are being retrieved from temporary memory and moved to working memory for processing. You will not remember every Type I dream; dreams without visual content may be difficult to remember.
How Memories Are Processed During REM Sleep
During REM sleep, your brain moves procedural memories from your temporary memory to your long-term memory. Your procedural working memory retrieves data from your procedural temporary memory, compares it with related files from your procedural long-term memory, and then deletes any unnecessary or duplicate data. The updated information is then transferred to your procedural long-term memory. The rapid eye movements that you experience during this time are a manifestation of this process.
You are almost completely paralyzed during REM sleep. Many people believe that the purpose of this paralysis is to prevent you from acting out your dreams so that you do not injure yourself. However, Zhang says that this cannot be the reason for paralysis during REM sleep, because you also dream during NREM sleep. According to Zhang, you become paralyzed during REM sleep so that you don’t physically respond to the procedural memories that your brain is processing.
Continual Activation
In order for your brain to work properly, both your declarative working memory and your procedural working memory have to remain constantly active. When the activity level of either the declarative working memory or the procedural working memory drops to a certain threshold, a continual-activation mechanism in the brain causes a random stream of memories to flow through that area of working memory. This keeps the activity level above the threshold.
According to Zhang’s theory, this continual activation process is what causes us to have vivid dreams.
Continual Activation During NREM Sleep
During NREM sleep, your declarative memories are being processed by your declarative working memory and then transferred to your declarative long-term memory. Your brain is focusing its activity on your declarative memories, not on your procedural memories. When the level of activity in your procedural working memory becomes too low, your brain starts to retrieve random data from your procedural temporary memory and procedural long-term memory and move it into your procedural working memory.
People usually don’t act out these random procedural memories because their muscle tones are reduced when they are asleep. However, sometimes the stream of memories may be too strong or the muscle tone may not be lowered enough. This can cause tooth grinding, sleep talking, sleep walking or periodic limb movement disorder, all of which are known to occur during NREM sleep.
Continual Activation During REM Sleep
During REM sleep, your procedural memories are being processed by your procedural working memory and then transferred to your procedural long-term memory. Your brain is focusing its activity on your procedural memories, not on your declarative memories. When the level of activity in your declarative working memory becomes too low, your brain starts to retrieve random data from your declarative temporary memory and declarative long-term memory and move it into your declarative working memory.
The first pieces of data that are retrieved will be easy, accessible information, such as day residue. These will cause Type I (thought-like) dreams.
As random memories continue to flow through your declarative working memory, your brain will try to make sense of them. Your associative thinking systems, which are located in the frontal lobes, and your emotion system will help to interpret these memories. This will cause you to have Type II dreams – more vivid, “dream-like” dreams.
The associative thinking systems strongly influence the plots of the Type II dreams. The associative systems add additional memories to the random memories that are retrieved during the continual activation of the declarative memory stores. They then use all of these memories to create realistic images and events and turn them into stories (dreams). This is why people with damage to their frontal lobes may not have vivid Type II dreams, even thought their continual activation mechanism still works (they do not fall into a coma).
Type II dreams mostly take place during REM sleep. However, you can have Type II dreams when you are experiencing NREM sleep and when you are awake.
Evidence
Zhang provides the following evidence for the Continual Activation theory:
We Don’t Remember Most of Our Dreams
We only remember our dreams if we recall them immediately after we wake up. This is because during sleep our temporary memory stores are in retrieval-only mode. Information can be transferred to working memory. However, working memory only has a limited capacity. In addition, the information stored in working memory is easily replaced if we are distracted by new information.
We Need Less Sleep as We Age
As we get older, the amount of time that we need for sleep decreases. While a newborn baby may require 16 hours of sleep a day, a 50 year old may require only six hours. The amount of sleep that someone needs depends on the amount of data they need to transfer into long-term memory, and the speed at which this transfer process occurs. Newborn babies have to file almost every memory they have into long-term memory
Newborns spend about half of their sleep time in REM sleep. This is because procedural memories are processed during REM sleep, and newborns have a great deal of procedural knowledge to absorb, as they are first learning the skills they need to control their body movements.
Newborns may also need more time to process memories since their brains are less mature. Adults with more well developed brains should be able to process memories more quickly.
H.M.
In this well-known case from 1953, a 27-year old patient known as H.M. had large sections of his medial temporal lobes removed. After his surgery, H. M. lost the ability to process new declarative memories. He was able to remember facts that he had learned long before the operation, which indicated that his long-term memory was not harmed. His procedural memory and his working memory were also unaffected.
Zhang’s memory model can explain the effect on H.M.’s memory. Zhang’s model says that there are separate memory stores for sensory, working, temporary and long-term memory, each of which is further divided into two separate stores, one for declarative memory and one for procedural memory. H.M. was unable to process new declarative memories because his declarative temporary memory store had been removed. All of his procedural memory stores, as well as his declarative working memory and declarative long-term memory were left intact.
Randy Gardner
In January 1965, San Diego high school student Randy Gartner earned a Guinness World Record for the longest amount of time spent without sleep. Randy stayed awake for 11 days (264 hours). Sleep researcher William Dement of Stanford University Medical School monitored Randy closely.
On his first day of recovery after achieving his record, Randy slept for 14 hours and 40 minutes. On the second day, he slept for 10 hours and 30 minutes. He slept for 9 hours on the third day.
Randy usually slept slightly less than 7 hours a day. Dr. Dement calculated that he therefore lost about 75 hours of sleep. However, Randy slept much less during his recovery.
According to the Continual Activation theory, the limited capacity of temporary memory explains why Randy did not spend more time sleeping during his recovery.
Randy had trouble staying awake by about the third day. This implies that temporary memory can hold only about three days worth of data. Therefore, after his first three days without sleep, Randy’s temporary memory was overloaded. He could save almost no new information to his temporary memory. The limited capacity of temporary memory explains why sleep deprivation makes it very difficult to learn and remember things.
The 14 hours and 40 minutes of sleep that Randy had on his first day of recovery, which is equivalent to slightly more than 2 nights sleep, would have cleared out most or Randy’s temporary memory so that he could process new data again. Zhang says that if Randy had only gone for three days without sleep, he still would have needed to sleep for about the same amount of time on his first recovery day.