PP15 The Real Philosophy of Science (and its implications for quantum weirdness)
G'day, my name is Bruce Robertson and this is pirate philosophy. In this series of videos I will be describing original philosophy, one that you won't find anywhere else but it is one that is logical, rigorous and dynamic. Welcome.
So far in this series of videos and our journey into philosophy, we have covered a variety of topics including the evolution of mind. The journey started with PP8'As Deep as You can Go In Philosophy', where we looked at the foundation of the separation of mind and matter. We then looked at how a logical processor can begin to create a model of the world and realised that the only logical process which could achieve this was one of assembling sense-data and then seeking to identify patterns within it. Subsequently, the identified patterns can be used as data in the Pattern-Identifying process in a recursive sort of way to create 'patterns of patterns' and continuing this recursive pattern-identifying process, a whole pyramid of patterns can be created, which then constitutes a model of the world.
If you haven't watched PP8, PP9 and Pp10, you might find it useful to do so before continuing this video. We also looked at 'reality and purpose', 'theory of consciousness', 'the logic of decision making' and in the previous video, we looked at mathematics and how it fits in with our model of the world.
Today, I want to look at science, i.e. the philosophy of science. So the title of this video is 'the real philosophy of science and its implications for quantum weirdness'.Some of the questions I want to answer, or at least a look at, are: 'what is the process of science?', 'Why is it been so successful?' and 'how can a philosophy of science contribute to science itself?'
Science is really just a continuation of the logical processes that we use to build a pyramid of patterns and a model of the world. But whereas much of the logical process for forming a normative model of the world (i.e. from basic sense-data) is hidden to our conscious minds, i.e. it is subconscious and we are not aware of it, science, on the other hand, is entirely explicit; It takes place in the higher realms of the pyramid of patterns, i.e. in the conscious mind.
And there is another difference as well, in that instead of receiving data in a fairly passive way and then making patterns or sense of it; science actively seeks out new data; such as doing experiments and developing specialized detectors that can reach beyond the realm of our simple senses. So, for example, the non-scientific normative view of the world looks for, and finds, simple patterns of the form that the Earth is flat, the Sun goes around the Earth and with regard to pushing a cart, a constant force produces a constant speed. But science has sought out data beyond the normative domain and has shown that there are better patterns that can be found that describe the same data and other data as well. Such as the Earth is a sphere, the Earth goes around the Sun and that when friction and air resistance are taken into account, a constant force produces a constant acceleration. In PP9: 'Patterns, Time and Space', I introduced an algorithm to show how information can be gleaned from sense-data with no prior assumptions as to the nature of the sense-data.
Here is that algorithm:
Simple Algorithm for Pattern Identification
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.
The essence of this algorithm is that it describes how first one assembles the data, in whatever form, primarily from sense-data. Secondly one inputs a test template this can be round, or a segment of the data or an identified person from an Antarctic different sets of data or whatever. then thirdly this complaint is tested to determine whether it fits the data -- i.e. that the template can be used to reproduce the data -- to a certain degree of accuracy. Fourthly, if the template fits the data it ends and stores the template together with a label for the data, if not its circles back to test another template. It is a fairly simple algorithm.
In this video, I want to discuss this algorithm in somewhat more detail and to see how it can apply to science.
So I have put together a more detailed (and complex) algorithm.
From a mathematical perspective, the aim of the algorithm is to compress the data. In order to do this one needs to find a pattern that can be described in a more compact form than the raw data and which also fits the data to a required degree of accuracy.
Perhaps the algorithm I'm describing applies most to physics, perhaps more so than other sciences but it certainly can be fitted to any science.
More Complex Algorithm
Pattern identifying algorithm:
1 Assemble the data, give it a label
2 Select a template (This can be random or a section of the data or one previously identified.)
3 Apply and expand the template to the data
4 Measure the differences between the applied pattern and the original data
5 Sum the differences
6 Store the template and the sum of the differences
7 Are there other templates to test? If ‘yes’ go to step 2, otherwise continue.
8 Of all the templates tested, select the one with the least sum of differences.
9 Store this template in association with the label for the original data.
So I'd now like to go over the algorithm, line by line:
Line one: Assemble the data and give it a label. The assembled data would need to have something in common with each of the other bits of data and also to the extent that it may be possible to find a pattern that describes all the data. The range of the data defines the domain of the pattern. The patterns have a limited domain, they fit the data but not necessarily can be extrapolated, you can extrapolate a pattern obviously but whether it fits in the domain, beyond the domain of the original data is something that has to be determined or checked.
Line 2: Select a template. This is perhaps the most interesting part of the algorithm, for it requires what we might call 'imagination' to concoct a possible template for the pattern. A template is a possible pattern; it is what will be tested to see if it fits the data. It can be anything at all; there is no limit to what could act as a template It could be entirely random or it could be a section of the data or it could be a template from an entirely different domain or it could be a combination of two or more other templates. Examples of templates in physics or in mathematical form might be 'y = x', 'y = 3x', 'y = +2', etc.
Line three: Apply and expand the template to the data. To show how this works I have put together some data, which is displayed in a two dimensional form.
Here, the data is a little more than a scatter diagram, but it will suffice for this example.
Then suppose that the first template is something like y equals x, which can be represented by a diagonal line.
So this diagram shows how the proposed pattern would look within the domain of the data.
Line 4: Measure the difference between the applied pattern and the original data. So what you can do here is to drop lines from the data points to the pattern as in the diagram.
In this way we can get an indication of the differences between the data and the pattern.
Line 5:Sum the differences. When you've got the differences, you can then add all the lengths of the differences to get an overall indication of how accurate the pattern is. This could possibly be refined by using the squares of the differences or using some other analytic method of assessing the accuracy of the pattern, but for this example, we will just add up all the lengths of the differences.
Line 6:Store the template and the sum of the differences. You need to store this information because we will be using it later.
Line 7:Are there any other templates to test? So here you can put in other templates, like y = 2x, or y = , or whatever you like and go through that process again, of applying the template to the data, looking at the differences, summing up the differences and then storing the sum of the differences alongside the details of the pattern. But if there are no more templates to test, then continue to line 8.
Line 8:Of all the templates tested select the one with the least sum of differences. So here we select the template or pattern that best fits the data. So in order to do that, you look through all the stored sums of differences to find the least sum. Then you take this template associated with the least differences and this is your best pattern.
Line 9: Store this template in association with the label for the original data. Store this pattern alongside the label given to the data in line one. So then you've got a label for the original data and you've got a pattern and you associate the two and you then store that.
This algorithm, or at least a variation of it, lies at the heart of the scientific process. I have shown a fairly generalized form of the algorithm, the bare bones if you like, of the scientific process. It can be modified so that it can be applied to any branch of science.
Also, I have shown this algorithm using only the criteria of accuracy to determine which pattern to select; but there is also the criteria of the simplicity of the pattern, which could be used if no one pattern stands out clearly. using the criteria of accuracy. (The simpler the pattern the greater the data compression)
Sometimes a particular pattern will fit the data exceptionally well and no other pattern fits the data at all well. In this case, the pattern can be used with confidence. For example, Newton's laws of motions are highly accurate in determining the motions of matter. In other domains, no particular pattern might stand out above the others. As for example, in trying to determine the connections and family tree of early hominids.
A pattern-identifying process such as this is the only way that knowledge of the world beyond our senses can be ascertained. As previously mentioned, it is a continuation of the pattern-identification process that we personally use to create our pyramids of patterns and model of the world.
Similarly to our own personal pyramid of patterns, science too has a pyramid of patterns, where one pattern is built upon the foundation of others. So for example, the theory or patterns relating to quarks are built upon many different patterns that have been established by conducting many different experiments that build up patterns relating to quantum particles, atoms, molecules matter and perhaps as far as our fundamental notions of time and space. So there is a pyramid, a hierarchy if you like, of patterns and patterns built upon patterns etc.
And in this way, our personal individual models of the world can be extended to new domains and new knowledge; and that is what science does.
Some may be concerned by the lack of certainty in the pattern- identification process, all that one arrives at is the best pattern and not the perfect one; or at least not one that can be shown to be perfect. But this is not a problem; as all that one needs for one's model of the world, whether personal or scientific, is sufficient accuracy to be able to interact effectively with the world. And science has shown this and done this with its technological spin offs, which have created things like cell phones, computers and other things that would be entirely impossible without an accurate model of the scientific world. But of course, this inherent uncertainty can create paradigm shifts within science as mooted by Thomas Kuhn; where a pattern within the pyramid of patterns that is fairly low down, gets replaced by a better pattern and this can cause a ripple effect on up to all the other patterns that relied upon it; this is a paradigm shift. So what one ends up with is an accurate model of the world, but one for which there may always be room for improvement.
What this means is that one has an accurate description of the world and while the laws of physics constitute an accurate description of how matter moves and interacts, one cannot claim, at least not within the philosophical domain, that matter obeys the laws of physics.
Perhaps I should also mention the role of mathematics in physics. Mathematics is an extremely useful tool for finding possible templates for the pattern-identifying process. However, as described in my previous video of PP14:'The Four Components of Mathematics', mathematics is, in essence, an entirely abstract system. Mathematics is an abstract system for the manipulation of symbols according to specified rules. And it requires a mapping from those symbols to elements within a pyramid of patterns for mathematics to have any connection to a non-abstract world. It requires a mapping between the symbols and a measurement or data or pattern, or whatever it is; it is a mapping between the two, and only in that way can mathematics relate to the world beyond our senses.
Science is a process carried out by many different people and groups of people and some may come up with different patterns or theories, which they claim to be good patterns or good theories but which may be at odds with other people's theories. So the question is then: 'What tests can one apply to a theory to determine its efficiency? 'And perhaps the best test is: 'Can the theory or pattern recreate the data to a sufficient degree of accuracy?' This test would be particularly effective in repudiating hand-waving pseudo-science theories which, in general, are entirely ineffectual at recreating data.
So that pretty much sums up the pattern paradigm approach to the philosophy of science. At the core of it is the algorithm I have described. And of course, once you have those patterns, you can extrapolate them and use them to create technology and all those sorts of things. 'And why an algorithm?', you might ask. I have used an algorithm because it shows explicitly the logical steps by which the process works. And while the algorithm may be crude and would certainly need some tweaking to fit each of the various scientific disciplines, it does demonstrate the essentials of the scientific method. And that scientific method is used for gaining information about the world beyond our senses.
And the reason that science has been so successful, at least compared to Standard Western Philosophy, is that it stays close to the facts of the world and does not venture into the domain of fantasy. Also its processes are a natural continuation of the way that we learn about the normative world of everyday life. In contrast, Standard Western Philosophy has attempted and failed to force fit the processes of science into its fantasy world of words and stubborn certainty. All that it has achieved, if you want to call it that, is to distort the essential nature of science.
Finally, I would like to look at some implications of this philosophy of science for science itself. I would like to go over a few suggestions or possibilities that might apply to the wonderful world of quantum weirdness. Patterns that are identified only have direct relevance to the domain of the data from which it was created or identified. While it may be extrapolated beyond this domain, it may not necessarily be accurate in the new domain. So, normative concepts of time and space and matter may not be effectively extrapolated into the domain of the very small or very large, or the domain of large masses and high velocities, as has been shown with special and general relativity and quantum mechanics and quantum weirdness. When one is trying to make sense of the unknown, as with quantum weirdness, no possible pattern or template should be rejected out of hand, simply because it is in contradiction with a pattern in a different domain. So it is not necessary for theories of physics to meet with normative perceptions of the world; they only need to fit the data and hopefully be able to reproduce the data. So there is no requirement to meet with dubious normative perceptions of existence, time, space, causality and so on.
The pattern-identifying process enables us both personally and in science to create a model of the world. This model resides in our brains, or minds, and it is quite distinct from the world beyond our senses whose existence we can only infer or hypothesize. In fact, the existence of a world beyond our senses is a best pattern to fit the relevant data.
It is perhaps naive, though understandable and acceptable for non-philosophers, to conflate the two and assume that our model of the world and the world beyond our senses are equivalent and identical. So all we actually know about the world is our model of the world and that can only be created through a process of pattern identification. In most circumstances it makes no difference whether you think that you have a model of the world or that you know exactly what is the world beyond our senses. But in the limits of science, this may become important.
And so this philosophical approach may be useful for resolving the famous Schrodinger cat thought experiment paradox. For while it may be the case that the best pattern to fit the data for a quantum particle is that it is in a superposition of states; i.e. that there are two states existing simultaneously, until further data is received or a measurement is made; this is only within the model of the world that we have. So, when it is claimed that Schrodinger's cat is in a superposition of states of being both alive and dead simultaneously, it is only our knowledge of the situation that is in a superposition of states and there is hence no requirement for the cat itself to be in a superposition of states.
One of the tests for a good pattern is that it can reproduce the data. Another requirement is that the pattern constitutes the compression of the original data, i.e. the pattern need to be simpler than the data and the more simple it is so much the better, for a given accuracy.
One of the proposed explanations for quantum weirdness and in particular, the unpredictability of the collapse of the wave function into a single data point is what is called 'Many Worlds'. This asserts that the wave does not collapse at all but that the world splits into many different worlds with a different outcome in each world. However, as a scientific theory, this 'Many Worlds' theory fails on both the criteria mentioned earlier. Firstly, it is incapable of reproducing the data of the world we actually live in and experience; it cannot reproduce the data of where the wave actually collapses. And secondly, neither is it a compression of the data. The Many Worlds hypothesis is highly complex, with many unanswered questions such as 'What happens in the other worlds?', 'Where are they?', 'How many other worlds are there?', 'What are the exact criteria for a world to divide into many worlds?' The complexity of the theory or pattern far exceeds that of the original data. In the light of this, it would seem that the Many Worlds theory or hypothesis is not a good pattern or theory.
Well, that is all I have for you today. If you have any interesting comments or questions about today's video, please leave them in the comments section below. And if you would like to continue this journey with me, then please subscribe to my channel, give it a thumbs up and ring the bell.
Thank you.
Comentários