No abstractions without representation

I should make it clear that Steve Grand and I agree on some very general things, but also disagree on a lot. I am much more extreme than Steve (something we both agree on), but Steve will be more than extreme enough for a lot of people.

I think the difference between "pattern instances" and "patterns" is largely semantic. I regard space-time as nothing special - as just another object - a pattern which exists because an algorithm applied to something else would, if applied, produce it. This means that, to me, an object existing at a particular point in space is merely a pattern described in very specific terms relative to another object: I just don't find spatio-temporal location profound or special and reject any philosophy that makes reference to space-time in any profound way if it is trying to say anything profound about ontology. I would say two electrons can exist in two different places because they are slightly different patterns, described by algorithms that are slightly different in terms of the parts of space-time they reference as "inputs." I would have no problem in describing patterns which are defined with reference to particular locations in space-time as "pattern instances," however: in fact, I already do that. The most recent article I wrote, which I have not posted on my website yet, and which is about AI and image analysis, makes extensive reference to pattern instances as a special case of patterns, for convenience.

Hi Paul, thanks for responding.

I think my main concern there was about the difference between a pattern that exists only in someone's mind, and one that is physically instantiated. Intuitively, of course, there's a big difference between these. But the more I think about it, the closer our positions seem to be. I'm going to take the liberty of doing a partial mind-dump here in the hope of getting some positive response. I'm beginning to think there's more chance of that here than some other forums I've tried it in.

Patterns and information are more-or-less the same thing, I think it's safe to say. However, I insist that there's two kinds of information. The first is very different from the ordinary concept of info, and many people find it very hard to swallow. The most generally accepted term for it, I believe, is "physical information". This is material form, things' shapes, but also all their other qualities, treated as Shannon information, and therefore quantifiable. This concept is quite well established in physics, consider for instance the conservation of information in black holes.

The ordinary concept of information, which unlike physical or Shannon information is meaningful, I call "intentional information". This is the concept of intentionality from philosophy of mind, so it's not about intention, but "aboutness": information in the usual sense of that word is intentional because it's always about something, even if what it's about is entirely fictional.

Now, I say that intentional information is always encoded in physical information, being decoded during use. This is nice, because it's a generalisation of Wittgenstein's later concept of meaning as use in context: you know the meaning of a word if you know how to use it in a given context. It's a generalisation because Wittgenstein's concept only covers language, or human communications at best, whereas mine encompasses all meaning and significance. A mind is a processor of intentional information. The content of consciousness is a stream of intentional information.

I need to say a little more about intentionality, though. I go along with Dennett, who says that it's not intrinsically a property of anything, but a concept we project onto things to make them easier to deal with: in particular, we view people as having thoughts, fears, beliefs, etc (about things), so we can (sometimes) predict their behaviour. That's why (or it's one of the reasons) I say that consciousness is ontologically subjective: whether an entity "has" a stream of intentional information "running through it" is a matter of interpretation. It suits us to view people -- and sometimes, to some extent, other species -- as conscious, but neither that, nor our conscious experience as individuals, makes consciousness an objective quality.

I could go on but that's probably enough for now. Any comments?

Hi Robin,

Sorry about the delay.

When I talk about patterns as things I'm not using the word in the same way as Paul is (as we've discovered through involved conversation). In fact I regret using the word at all. "Form" is closer than "pattern" to what I mean, but that's too Platonic.

All I mean is that an instance of a "phenomenon" - whether it be a tornado or a creature or an atom - is really a process. It's a self-maintaining disturbance in something, and the spatiotemporal form of that disturbance is key to its properties, including the mechanism that causes it to persist and resist disruption. This spatiotemporal form is what I've been calling a "pattern" for want of a better word. So it's not like a checkerboard pattern or even a chair-shaped pattern, which I agree with you is a product of the mind. The discussion with Paul has wandered a bit from where it started - he is talking about different things to me.

I wish I could think of a better word. "Arrangement" is better but too weak, since everything has an arrangement, even if it serves no purpose. A good analogy for what I'm talking about is an electronic circuit: it is the arrangement of its parts that creates the circuit's properties and thus gives it an identity. You can build a radio receiver, a radio transmitter and an oscillator from exactly the same components; it's their arrangement (the network of interactions between the properties of the components) that defines their existence as distinct entities.

A living thing is an arrangement of chemical reactions. It is a process. It is the arrangement of this network that is alive, not the atoms that it is made from, since these atoms are constantly being replaced. The body is in flux but the pattern - the arrangement - persists. It's the circuit of interactions between properties that gives the network persistence.

And I assert that even an electron or an atom is a process - an arrangement of something. Solid matter is really more like a whirlpool in water than a lump of something sitting on space (the demise of the Aether theory notwithstanding). Atoms are resonant states. Look at the shapes and energy relationships of electron orbitals - they are more like the patterns of vibration on a ringing bell than little planets whirling round a star. They are self-maintaining disturbances in the electromagnetic field, and it is the "shape" of these disturbances that makes them able to persist and interact with other disturbances (i.e. have properties).

So the "patterns" I've been talking about are rather specialised compared to the sort Paul has been talking about.

Hope that helps clarify things - I doubt it, because I'm pretty confused myself at the moment...

- Steve

I thought I should try to put in an appearance again.

Steve and I are probably talking about different things because we think that different things are needed for "reality". I want to say here that I do not really like the word "pattern" either - with regard to my own views - but for different reasons than Steve. Nevertheless, I have used the word "pattern" because it was the best word I could think of.

I don't like the word "pattern" because it suggests that I think some spatial arrangement of things is special - it suggests that I have some kind of interest in geometry. I can imagine some kind of geometry-obsessed caricature of me, buying lots of "join the dots" children's books and joining the dots in weird ways and screaming that the results are all real. This would not be totally accurate, because I regard space itself as just a special case. My view of a "pattern" is that it is some algorithm that can be applied to an object to generate another object and that is it. Objects are data. Patterns are algorithms. Patterns generate data. I am not saying that anything has to "run" these algorithms. I am simply saying that when we can run one of these algorithms and produce an object, that algorithm is describing something that exists in reality. I regard the distinction between pattern instances and patterns as largely semantic - in that a pattern instance is merely a particular pattern which is a special case of a more general pattern. Does this mean I reject what Steve says about emergent properties? Not at all. I actually agree with it. I too think that an electron is a process. I think everything we see is a process - some kind stable self propagating pattern in space-time. I think he describes very well particular instances of patterns which are going to most relevant to us, given our place here, in space-time. However, I regard everything around us as a special case, and think that what Steve is saying applies very well to this special case. My own view is that basic concepts of ontology should not make reference to space and time because space and time are just things that we empirically observe. In other words, I think Steve and I agree on what is happening, here, for the class of things we observe, what the relationship of those things to space-time is, what the things that we see "existing" or doing, etc. If Steve and I look at an atom, a whirlpool or a brain we are going to see the same things and the same kind of ontology (more or less). The difference between Steve and me (I think) is that I assign the status of "real" to a lot of extra stuff that is not covered by all this. Another way of looking at this is that Steve (I think) regards processes as critical in giving rise to "objects" while I regard the idea of "process" itself as a special case that is only relevant here and anywhere else where you happen to have a space-time - or at least a time.

BTW Steve: I posted a scenario involving mind uploading in the "main debate" and I will try to answer the other points when I can. Sorry about the delay.

I disagree with Dennett on this issue. I regard things like intentions, emotions as real because I have them and because they are reasonable concepts to make sense of what matter is using. I do not think there is anything mystical in this. Particular classes of processes can be labelled as "emotions". Other processes can be labelled as intentions. I do not see why it helps to give any of this the status of "unreal". if we are going to do that what would stop us declaring any object unreal on the basis that nothing is there except some process - or a self propagating pattern - which we interpret as "a real thing" for convenience. As an example, it could be argued that tornadoes do not really exist: it is just convenient to use the idea to describe what is going on when houses and cows get sucked up into the air. I choose to say that things are real. If we are going to say that my emotions are real - given that my emotions are a particular pattern, or process, in my brain, what makes them any less real than all the other things we see? Could we not just say that reality is a mess of atoms and it is "convenient" for us to say that certain objects exist in that mess? In fact, how can we even say a table is real, when all we know about that table is as object being recognised by a mind that is itself not supposed to be real?

Trying to put this another way, it is all very well to say that we only use words like intention to try to make sense of what other people's brains are doing, but that is all that any words are for. We use words like "table" to describe what table atoms are doing. We use words like "giraffe" to describe what giraffe atoms are doing. We use words like "atom" to describe what subatomic particles are doing. It seems strange to me that Dennett seems to make some kind of special case when brains are involved. Would we say that computers are not really running Windows, that it is just a convenient fiction to help us understand how our computers are going to behave? I would say that "Windows" describes a real arrangement of electrons – or more accurately, real processes involving electrons. Every computer running Windows has objective properties in common.

If the mind is an illusion, it raises the question: who is supposed to be getting fooled by it? This may seem a trick, as when theists ask, "Who made the universe?" when trying to prove that someone made it, but I think it is more valid here: the whole idea of an illusion is that someone is being fooled.

Hi Steve, sorry for the much longer delay! I was under the impression that I'd get an email when anybody posted to this thread, and was wondering what you guys were doing, but now I know!

I really think "processes" is the word you're looking for there. It has both the dynamic and the open-ended/fuzzy bordered implications you're looking for, I believe. Otherwise, I see things very much as you do!

Sorry for the delay, for my excuse see the reply I just did to Steve.

You say you disagree with Dennett but I'm not sure whether you really do or not. It might just be that the view he's trying to deny is one you don't take seriously enough to bother with at all. The question is not so much whether intentionality is "real", but whether it denotes anything special. In fact, "reality" is an extremely slippery concept that I'm slightly nervous to see bandied about so freely here. What does it "really" mean?? Well, let's not get into that now!

The thing about intentionality is that it "refers" to something outside itself. So what does "refer" mean? Dennett, I'm fairly confident, would say that the connection between a person and, for instance, what they're currently looking at, is just a matter of contingency, not a matter of principle or logical necessity. It just so happens that light bouncing off that thing enters the eyes and, following a lot of processing at various levels, somehow corresponds with a "mental representation", whatever that is, such that the person can reach out and pick the thing up without missing it or knocking it over. OK, so "mental representation" is quite mysterious, but what it hides is supposed, by Dennett and me, to be nothing metaphysically special, or specially metaphysical. Obviously, this intentionality thing works for us, and so can be considered real if you like, but it's not a thing-in-itself, just a temporary arrangement of other things brought together for a particular purpose, which might or might not work. Maybe you don't think anything is a thing-in-itself, so don't see what the problem is, in which case you're at odds with many, many people, but not with Dennett or me, except that we see the need to communicate with these others, which you perhaps don't. Have I got that right?

I think things like mental states, intentions, emotions, etc exist. I do not view them as anything special, in any profound, ontological sense: of course they are special to us because our life revolves around such things, but that is just personal.

My understanding of Dennett was that he had actually stated that consciousness was not real. Now, I would disagree with that. On the other hand, if someone told me that consciousness was real, and was just a "thing" - in the same way that trees and rocks are things, I would agree. I did think that Dennett went a bit further, though, and actually declared things like consciousness and intentionality not to be real. My view of intentionality is simple, ontologically: it is a load of neurons doing those things that we label as "having intentionality".

As far as I am concerned, something like intentionality exists by being an emergent property of groups of neurons, which are in turn an emergent property of groups of atoms and so on. I certainly do not think there is anything metaphysical associated with it. The fact that my consciousness can be reduced to a low level description in terms of arrangement of matter, to me, does not detract from its existence at all, any more than doing the same to a tree would detract from its existence.

It may be that all this is down to semantics. Dennett and I seem to agree on what is going on at a low level, and how that is causing what we observe at a low level. It may be down to different understandings of words like "illusion". I really don't like this word "illusion", because it suggests that what we experience is not real. If, on the other hand, it is merely suggesting that what we experience has lower level causes then I have no problem with it.

Maybe Dennet, I and you just use words a bit differently?

Regarding the issue of whether I think anything is a thing in itself, I would need to ask what the "in itself" means. My first instinct, however, is to say "probably not" - I think that a thing is a thing because it is some abstraction of other things. I do not distinguish between "abstract things" and "things that are real". One way of looking at this would be to say that I do not recognize abstraction as some kind of logical process that constructs "abstract things": I just regard it as part of the ontology of things in general. So is a thing ever a thing in itself? Probably, no.

Paul said "Maybe Dennet, I and you just use words a bit differently?"

I think that's basically right, but I don't think Dennett ever said consciousness isn't real. That's what some people interpret him as saying, but that's a different thing. I can't remember the details -- it might be in The Mind's I -- but he has said something to the effect that "we obviously do have all these experiences", which to me implies, at least, he wouldn't view consciousness as unreal. The self, on the other hand, he does view as an illusion, though a benign one. And I agree with him on that as well. It doesn't mean that any experience is illusory, because we don't experience the self, it's a concept drawn from experience, an interpretation. What do you think?

I think I would have to take another look at what Dennett said before I can comment further.