top of page
Search
outml99

PP10 Pyramids, Patterns and Why we Sleep


G'day, my name is Bruce Robertson, and this is Pirate Philosophy.


In this series of videos, I will be describing an original philosophy, one that you won't find anywhere else, but it is one that is logical, rigorous and dynamic. Welcome.


In this video, we are continuing our journey, which started with the postulates of a logical processor with senses and motor outputs.


In the previous video "PP9 Patterns Time and Space", we discussed how the logical processes of pattern identification could be used to begin to make sense of raw sense data. And at the beginning we talked about a very primitive logical processor as might have been the case at the very start of animal evolution. But now our animal has thrived because of its ability to interact effectively with the world and a few hundred million years or generations later, it has evolved a larger brain which is better able to make sense of the world. It was also noted that the process of pattern identification is the only logical process by which raw sense-data can be interpreted without any form of preconception of what the raw sense-data might comprise. The algorithm for this logical process being something like:


1 Assemble the data

2 Input a template

3 Test the template – does it fit?

4 If it fits continue otherwise return to step 2

5 Store the template together with a label for the data.

While this process is able to simplify and begin to make sense of the world from the raw sense-data, something more is required to create a comprehensive model of the world such as we humans experience. So what is it?

Let us look at some examples of the sort of patterns that might be created from sense-data. And for these examples I will be using simple mathematics to model the examples.


Data Pattern Label

1 2 3 4 5 n A

2 4 6 8 10 2n B

3 6 9 12 12 3n C

4 8 12 16 20 4n D


In these examples 'Data' represents the sense data, 'Pattern' represents the templates or pattern that best fits the data and 'Label' is just a label to identify that particular pattern, and 'n' denotes the position of a data element in the sequence.


So, suppose your data took the form of 1, 2, 3, 4, 5 then the pattern would be 'n' where 'n' denotes the position of the different data elements, and then we might label that pattern as A.


Then suppose that we had some more data which took the form 2 4 6 8 and 10. Fairly obviously the pattern would be 2n and we can label that as B. Some more data 3 6 9 12 and 15; the pattern identified then is then 3n and we can label that C. And we have some more data 4 8 12 16 and 20 so fairly clearly the pattern is 4n and we might label that D. (Note that the identifying pattern is a compressed form of the data and that by using the pattern we can recreate the data.)


So now we have some patterns that fit the raw data where do we go from here? What if now instead of using the raw sense data as the input to the pattern identifying process we used the identified patterns as input. Then we would have as our input data n, 2n, 3n and 4n.


Data Pattern Label

1 2 3 4 5 n A

2 4 6 8 10 2n B

3 6 9 12 12 3n C

4 8 12 16 20 4n D


n 2n 3n 4n m X n ZZ


And if we denote the position of the data by the letter m then the pattern would be m times n. we might label that ZZ.


What we have arrived at is a pattern of patterns.

And this process of putting the identified patterns back into the pattern identifying process as input data does not have to stop there.


When a sufficient number of patterns of patterns have been identified at what might be considered a level two pattern, these can be assembled and subsequently used as input to the pattern identifying algorithm to generate a level three pattern.




Data level 1 Patterns level 2 patterns

1 2 3 4 5 n

2 4 6 8 10 2n m X n

3 6 9 12 12 3n

4 8 12 16 20 4n


And this process of taking the output of a logical process and reinserting it as input for the same logical process is called recursion. And because pattern identification is necessarily the compression of the data the quantity of patterns in each level will be smaller than the level before it so then if we draw a schematic diagram for these levels of patterns we arrived at something that has the shape of a pyramid or more accurately a step pyramid where each higher level being smaller than the level below it. So then you have your sense data at the bottom, which is just your input raw data, then you have your level one patterns, level two patterns, level three patterns and so on up to level four or however far you want to go. And in this way we create a model of the world based on raw sense data.


So let us look at another example of how this might work. For this example, I will only be using the labels of the patterns to designate the patterns as the actual details of the patterns themselves would be highly complex. Nevertheless, it is actually the details of the patterns which are used in the pattern identifying process and these labels well, they’re just labels.


Suppose at level one, one has the labels for the patterns of 'leaf', 'branch', 'fruit' and 'trunk':




then these might be combined into a pattern which is then given the label tree then this level 2 pattern of tree could be combined with other level 2 patterns such as 'bush', 'flower' and 'vine'.



and then together with tree these might be combined into a pattern found which is then given the label 'forest'.


and that would be a level 3 pattern then this level 3 labeled 'forest' can be combined with other level 3 patterns which might be 'grassland', 'river' and 'hill'.



Then they might be combined into a pattern which is given the label 'countryside'.


This would be a level 4 pattern and the level 4 pattern labeled as 'countryside' could be combined with other patterns for things such as 'town' or 'sea'


In this way, one arrives at a pyramid of patterns. And this pyramid of patterns constitutes one's beliefs or knowledge about the world. It constitutes a model of the world.


Only through the pattern identifying process, can a model of the world be created.


The actual process of pattern identification through all these levels is a very slow and lengthy one. Data has to be assembled, possible templates tested then the accuracy of the result pattern evaluated before a final pattern is selected.


It is a long process and so for higher patterns to be found this would be best done when the logical processor or brain is not busy doing anything else like evaluating raw data from its senses searching for food, invading predators and so on. So, this would best be achieved when the animal has found a place of safety and closed off its senses so that its logical processor or brain can focus on searching for higher level patterns or in other words when the animal is asleep.


So, this provides an explanation for why we need sleep. It is an essential requirement for our brains to create a comprehensive model of the world and it is also an explanation for why babies need so much sleep as their brains work overtime in searching for patterns that will enable them to make sense of the world in which they find themselves. And it should be noted that this explanation for sleep is emergent from the theory we have been discussing; from the assumptions and logic already described in this series of videos. And the fact that it is emergent and that it fits well with our experience is further evidence that we are on the right track on this journey.


In standard western philosophy, there is no philosophy of sleep. Even psychologists have no explanation for it. So the only people who know why we sleep and the logic of what our brains are doing when we sleep are you, me and other viewers of this channel.


Finally in this video, I would like to discuss very briefly the labels that are used to identify patterns.Patterns that are near the base in the pyramid of patterns have labels that remain hidden to us. So, for example, Low-level patterns, (low in the pyramid of patterns that is), have labels that remain hidden to us. So, for example, patterns that we use to interpret light and shade as shadows are hidden from us and yet they enable us to perceive a three-dimensional world from the two-dimensional data we receive from our eyes.


However, for higher level patterns the labels are equivalent to what we call 'words'; or to put it another way words are labels for patterns. Words like 'run', 'tree', 'elephant' and so on are all labels for patterns that have been created ultimately from sense data through the pattern identification process.


It is interesting to note that Standard Western Philosophy has no theory for what a word is. Here is a quote from something I found on the Stanford Encyclopedia of philosophy on the internet: “The notions of word and word meaning are problematic to pin down, and this is reflected in the difficulties one encounters in defining the basic terminology of lexical semantics” or in other words they have no idea what a word is. And yet their whole philosophy is based on words and the manipulation of words; all without any viable theory of what a word actually is. But as we have seen and according to The Pattern Paradigm, which I am describing in these videos, a word is a label for a pattern. So it seems again that the only people who actually know what a word is are you, me and other viewers of this channel.


Well, that is all I have for you today. If you have any interesting comments of questions about today’s video, please leave them in the comment section below. And if you would like to continue this journey with me then please subscribe to the channel, give it a thumbs up and ring the bell.


Thank you.













13 views0 comments

Recent Posts

See All

Comentarios


bottom of page