Neurologist outlines why machines can’t think

[Silmarien]

So how do we distinguish 'inherently goal-directed' behaviours from non-inherently goal-directed behaviours, what are the distinguishing criteria?

Perhaps we don't. Perhaps all of nature is inherently goal-directed. Though keep in mind that goal-directed doesn't necessarily entail conscious behavior--the chemical reactions that move an amoeba towards its food aren't an accident; they develop and are carried out specifically because the amoeba needs to eat. Where should we start with the chain of causality: the goal or the chemical reactions?

I'm fine with calling them different levels of description, but can we reduce one to the other? That seems to be what you're trying to do, and I'm unconvinced that you can avoid absurdities like the elimination of rationality or magical inexplicable jumps in functioning à la Intelligent Design with this approach.

We could say that two chemicals 'want' to react together in order to generate a product because one has electrons to share and the conditions are right, but when the 'reason' or 'motivation' obviously reduces to thermodynamics, it's hard to attribute agency... we generally reserve that kind of description for children, who seem to find it easier to understand the world in terms of agency and teleology.

Well, assuming that the motivation obviously reduces to thermodynamics. This depends upon what laws of nature actually are--if the camp that thinks they're descriptive rather than prescriptive is correct, then nothing actually reduces to thermodynamics. Thermodynamics is simply the way we're describing the interactions of physical substances that are in fact reacting based on their intrinsic natures.

The evolutionary explanation seems quite reasonable - creatures that tend to avoid dangerous or damaging stimuli have a selective advantage; multicellular creatures evolved nervous systems to coordinate their responses to environmental stimuli.

It may seem reasonable, but it's actually deeply flawed. Here's the problem:

Let's take two hypothetical amoebas. Amoeba 1 is a chemically driven specimen with no experience of hunger. Amoeba 2, in contrast, feels hunger.

What is the difference ontologically between Amoebas 1 and 2? This is the underlying question. In step one of our thought experiment, let us assume that they are chemically identical, but that Amoeba 2 mysteriously possesses sensation as well. If they are physically identical specimens, then is Amoeba 2's mysterious experience of hunger going to give it a selective advantage? What causal role is this strange Cartesian sensation going to be playing in Amoeba 2's eating habits? If behavior is determined at the physical level, then there is no effective difference between Amoebas 1 and 2. For the purposes of evolutionary processes, they are identical.

Now, perhaps Amoebas 1 and 2 cannot be identical. Perhaps additional chemical reactions are required for Amoeba 2 to feel hunger, so the physical make-up of the two amoebas must be different. So we can now posit that Amoeba 2 is somewhat different than Amoeba 1 physically, and that it is this physical difference that makes it more fit for survival. All well and good, but enter Amoeba 3: another amoeba that is physically identical to Amoeba 2, except without phenomenal experience. Once again, it is as likely to thrive as Amoeba 2, since all causality is occurring at the physical level, and we're off on an infinite regress of unhungry amoebas.

If the phenomenal can be reduced to the physical, then evolution cannot be invoked to explain the presence of phenomenal experience. If it is something additional to the physical that can be selected for by evolution, then we're into a form of dualism and worrying about immaterial causation.

Of course, we could also deny that Amoeba 1 is possible and say that at a certain level of development, the amoeba will automatically be hungry. But why would this be the case? It's not a development that we can attribute to evolution, for the reasons listed above. If there is a logical reason why physical systems are accompanied by phenomenal experience, it's going to go deeper than evolutionary advantage, and it's almost certainly going to land us outside of a strictly physicalist ontology.

I don't think anyone has satisfactory explanations for consciousness, qualia, etc. That's the 'hard problem'. I currently see it as, in Nagel's terms, that there is just 'something it is like' to be a creature doing this kind of information processing.

That's fine, but Nagel is a serious opponent of materialism, so I would worry about the coherency of a Nagelian flavored physicalism.

My problem with the 4D 'block' universe is why, if it all just is in a lump, should it show a spatial progression in the temporal dimension; i.e. why is each temporal 'slice' related to the previous and subsequent slices such that entropy appears to increase from slice to slice?

Ahh, yes. I have a somewhat similar problem with the Block Universe theory, if more observer based. There is no change, change is an illusion, but we will refuse to address the fact that your subjective experience of change is still an actual change that we need to account for. Enter lots of hand waving.

 

[FrumiousBandersnatch]

Perhaps we don't. Perhaps all of nature is inherently goal-directed.

If you like - but as I said previously, it doesn't seem a particularly useful interpretation for low-level phenomena. YMMV.

Though keep in mind that goal-directed doesn't necessarily entail conscious behavior--the chemical reactions that move an amoeba towards its food aren't an accident; they develop and are carried out specifically because the amoeba needs to eat. Where should we start with the chain of causality: the goal or the chemical reactions?

The goal is just a high-level way of describing the effects of the behaviours emergent from the chemical reactions. Either is valid, their utility depends on what you're trying to do. If you're describing how an amoeba fills its environmental niche, the high-level description may appropriate; if you're describing how the amoeba moves, feeds, and regulates its internal environment, the chemical description may be more appropriate.

I'm fine with calling them different levels of description, but can we reduce one to the other? That seems to be what you're trying to do...

When the higher level is emergent from the lower (which seems to be the usual situation) we can reduce one to the other in terms of describing how a specific event or concept at the higher level is the result of certain interactions at the lower level, but we can't generalise this relation, i.e. we can't predict behaviour from chemistry or chemistry from behaviour - unless we've already traced the detailed correspondences.

Some examples are clearer than others - in Conway's Game of Life, we know the low-level details and simple rules of the automaton, but they tell us nothing about the ways dynamic patterns on the grid move and interact. In fact, a full description of the grid and its rules tells us nothing about what will happen when it is iterated over; that depends entirely on how the cell states are initialised.

Well, assuming that the motivation obviously reduces to thermodynamics. This depends upon what laws of nature actually are--if the camp that thinks they're descriptive rather than prescriptive is correct, then nothing actually reduces to thermodynamics. Thermodynamics is simply the way we're describing the interactions of physical substances that are in fact reacting based on their intrinsic natures.

True enough - although if thermodynamics is simply the way we're describing the interactions, then they clearly do all reduce to thermodynamics - but what that means depends on your interpretation of the interactions thermodynamics describes.

My point is that these are our interpretations of what we observe; things interact and we can choose to interpret those interactions in mathematical terms (e.g, physics, chemistry, etc.), or teleologically - which seems to me to be an anthropomorphic projection at the lower levels. The question is how useful each form of description is at various levels of application, and the implications of that application.

Given that, as I mentioned previously, teleology is strongly associated with agency - presumably because agents act according to internal state changes that are opaque to external observers, and are therefore somewhat unpredictable, which is why we tend to describe their behaviour in terms of rather vague teleological abstractions - it makes sense to use teleological descriptions for agent behaviour.

However, interactions at the molecular level are relatively transparent and amenable to precise and repeatable calculation (albeit sometimes probabilistic), and as thermodynamics is a precise and repeatable statistical treatment of atoms & molecules in bulk, it seems reasonable to me to treat behaviours at that level as non-teleological, maintaining the utility of the teleological description as a means of distinguishing the opaque, high-level behaviours of complex agents from the transparent, highly predictable low-level interactions of simpler systems.

It may seem reasonable, but it's actually deeply flawed. Here's the problem:

Let's take two hypothetical amoebas. Amoeba 1 is a chemically driven specimen with no experience of hunger. Amoeba 2, in contrast, feels hunger.

What is the difference ontologically between Amoebas 1 and 2? This is the underlying question. In step one of our thought experiment, let us assume that they are chemically identical, but that Amoeba 2 mysteriously possesses sensation as well. If they are physically identical specimens, then is Amoeba 2's mysterious experience of hunger going to give it a selective advantage? What causal role is this strange Cartesian sensation going to be playing in Amoeba 2's eating habits? If behavior is determined at the physical level, then there is no effective difference between Amoebas 1 and 2. For the purposes of evolutionary processes, they are identical.

Ah... this misses the point of my using an amoeba as an example. I apologise for not making it clearer; the idea is that by using an amoeba as an example, where we can conceive of its food-related behaviours being governed by relatively simple chemical sequences and feedbacks, i.e. simple mechanical reflexes, nevertheless we can describe its behaviour in high-level anthropomorphic and teleological terms, e.g. hunger, seeking food, like & dislike, satiety, etc. So I'm not suggesting that the amoeba has subjective sensation (for this hypothetical, it doesn't), but that we can describe its behaviour in those terms (the intentional stance), and it is tempting to do so. And I think we have a tendency to conflate or equivocate our own unconscious behaviours with conscious deliberative activity, using the intentional stance and confabulation (generating a plausible narrative to explain non-deliberative activity).

That's fine, but Nagel is a serious opponent of materialism, so I would worry about the coherency of a Nagelian flavored physicalism.

I think it's fine to pick and mix concepts and terminology from various sources - it's the way they're assembled that makes philosophy. In this case, it's a simple expression that captures the concept that certain arrangements of self-referential mapping and monitoring can give rise to that self-sense; i.e. that there is 'something that it is like' to be that system.

Ahh, yes. I have a somewhat similar problem with the Block Universe theory, if more observer based. There is no change, change is an illusion, but we will refuse to address the fact that your subjective experience of change is still an actual change that we need to account for. Enter lots of hand waving.

In Tegmark's description it's the flow of time that's illusory. Change is the variation between slices on the time axis (i.e. along world-lines). The assumption is that subjective experience is a particular kind of information processing, and is encoded in the changes from slice to slice as 'observer moments'; each moment is a snapshot of you 'mid-thought', and any sequence along your world-line is you having some subjective experience. He makes the comparison with a reel of film or a movie DVD, but there is the significant difference that a film or DVD needs a player to iterate over successive frames (because the observer is not part of the movie), and there is only a single progressive 'now' in any showing, consisting of the frame 'in the gate' at some time; this confuses things.

 

[Silmarien]

True enough - although if thermodynamics is simply the way we're describing the interactions, then they clearly do all reduce to thermodynamics - but what that means depends on your interpretation of the interactions thermodynamics describes.

I am worried about your use of the word "reduce" here. If thermodynamics is the result of physical substances interacting acording to their intrinsic natures, then what we see as the law of thermodynamics is actually an abstraction, not a fundamental aspect of reality. If reality can be reduced to abstractions, then you have chosen your username well, because this conversation just took a sharp turn into Wonderland.

Given that, as I mentioned previously, teleology is strongly associated with agency - presumably because agents act according to internal state changes that are opaque to external observers, and are therefore somewhat unpredictable, which is why we tend to describe their behaviour in terms of rather vague teleological abstractions - it makes sense to use teleological descriptions for agent behaviour.

Teleology is associated with agency amongst opponents of teleology, and of course amongst the general public. If you actually look at the Neo-Aristotelians who invoke immanent teleology as a principle of nature, the first thing they do is deny that teleology has anything to do with agency at all.

Ah... this misses the point of my using an amoeba as an example. I apologise for not making it clearer

No worries; you were clear. I apparently was not, since I was using the amoeba example to illustrate a separate point--the problem with invoking evolution to explain phenomenal sensations in general. It may not turn up with amoebas, but it will show up at some point along the evolutionary chain. If you are going to hold that all causation is taking place at the physical level, then evolution cannot select for any aspect of phenomenal experience. Pain offers no evolutionary advantage if it is merely an abstraction of physical processes. It is those processes that offer the advantage, and the fact that they are accompanied by sensory pain still needs to be accounted for. (Or accepted as a brute fact, I suppose, though I tend to take brute facts as admissions of defeat.)

One major problem here is dialectical imprecision. You have been referring to experiential phenomena as "abstractions" without first pinning down what is meant by the word in this context. Because of this sort of careless use of language, we've been slithering around between various mutually exclusive versions of materialism at will, picking whichever interpretation is most suitable for a specific situation while some sort of materialistic chimera which is simultaneously eliminative, reductive, and emergent takes form around us.

I think it's fine to pick and mix concepts and terminology from various sources - it's the way they're assembled that makes philosophy. In this case, it's a simple expression that captures the concept that certain arrangements of self-referential mapping and monitoring can give rise to that self-sense; i.e. that there is 'something that it is like' to be that system.

Very well. How does this happen? How do certain arrangements of purely physical self-referential mapping and monitoring give rise to a self-sense? This is where I'm most interested in seeing a plausible causal history.

 

[FrumiousBandersnatch]

I am worried about your use of the word "reduce" here. If thermodynamics is the result of physical substances interacting acording to their intrinsic natures, then what we see as the law of thermodynamics is actually an abstraction, not a fundamental aspect of reality. If reality can be reduced to abstractions, then you have chosen your username well, because this conversation just took a sharp turn into Wonderland.

Of course thermodynamics is an abstraction - it's a statistical description of the behaviour of vast numbers of atoms and molecules.

Having said that, all our descriptions of reality are abstractions. That doesn't mean reality itself is an abstraction, but we can only describe it in terms of shared concepts using language. Even our personal experiences and perceptions are abstractions - we infer or construct an internal representation of reality from the signal spike trains entering our brains from our sensory nerves.

Teleology is associated with agency amongst opponents of teleology, and of course amongst the general public. If you actually look at the Neo-Aristotelians who invoke immanent teleology as a principle of nature, the first thing they do is deny that teleology has anything to do with agency at all.

I don't have a problem with that, there are many different philosophies (although it seems possible to attribute Aristotelian final causes to the agency of nature as a whole, or to the agency of the interacting elements individually - either way, it seems an idealist anthropomorphisation) - I was suggesting a plausible explanation for the general tendency to associate teleology with agency.

No worries; you were clear. I apparently was not, since I was using the amoeba example to illustrate a separate point--the problem with invoking evolution to explain phenomenal sensations in general. It may not turn up with amoebas, but it will show up at some point along the evolutionary chain. If you are going to hold that all causation is taking place at the physical level, then evolution cannot select for any aspect of phenomenal experience. Pain offers no evolutionary advantage if it is merely an abstraction of physical processes. It is those processes that offer the advantage, and the fact that they are accompanied by sensory pain still needs to be accounted for. (Or accepted as a brute fact, I suppose, though I tend to take brute facts as admissions of defeat.)

To answer that one has to account for phenomenal experience, which is a complex and speculative subject; but Damasio's neurologically-based account of the how the (evolutionarily) early mappings of the internal milieu to a dynamic body image in the brain, gave rise to primordial feelings ('Self Comes to Mind' Ch.4 'The Body in Mind', p.89), which subsequent evolution elaborated in richness and sophistication, and yet the primitives of pain and pleasure continued to be as powerful and dominating as ever, suggest that they are extremely effective and successful behavioural drivers.

One major problem here is dialectical imprecision. You have been referring to experiential phenomena as "abstractions" without first pinning down what is meant by the word in this context. Because of this sort of careless use of language, we've been slithering around between various mutually exclusive versions of materialism at will, picking whichever interpretation is most suitable for a specific situation while some sort of materialistic chimera which is simultaneously eliminative, reductive, and emergent takes form around us.

By abstraction, I typically mean a generalisation or indirection of properties or relationships (my usage is probably influenced by my career in object-based software design, where abstract types are ideals or templates that cannot be directly instantiated, and to which concrete instantiable types have an 'is-a' relationship; e.g. an apple is a fruit). Most people don't seem to have difficulty understanding what I mean, but if you have trouble following, just ask and I'll try to explain.

How does this happen? How do certain arrangements of purely physical self-referential mapping and monitoring give rise to a self-sense? This is where I'm most interested in seeing a plausible causal history.

I don't know, nor, it seems, does anyone else; but they do. The main difficulty is the fundamental perspectival divide between objective description and subjective experience; however many objective correlates of self-awareness or consciousness we are able to establish, we have no access to the nature of subjective experience other than our own personal experience - and even that is evanescent.

 

[Silmarien]

Of course thermodynamics is an abstraction - it's a statistical description of the behaviour of vast numbers of atoms and molecules.

Having said that, all our descriptions of reality are abstractions. That doesn't mean reality itself is an abstraction, but we can only describe it in terms of shared concepts using language. Even our personal experiences and perceptions are abstractions - we infer or construct an internal representation of reality from the signal spike trains entering our brains from our sensory nerves.

Wait a moment, are you a transcendental idealist who just prefers the language of materialism as some sort of constructed subjective reality? Because this is what I mean when I say that your position is self-contradictory--you cannot simultaneously hold that objective reality is unknowable in the Kantian sense and hold to metaphysical materialism. (Except as an object of transcendental faith, but that would be a pretty strange move on an ontology that denies transcendence.)

I don't have a problem with that, there are many different philosophies (although it seems possible to attribute Aristotelian final causes to the agency of nature as a whole, or to the agency of the interacting elements individually - either way, it seems an idealist anthropomorphisation) - I was suggesting a plausible explanation for the general tendency to associate teleology with agency.

Every attempt to superimpose human concepts upon reality is going to be anthropomorphism. This is as true for those who view nature through the language of machinery (or in modern computational terms) as it is for older paradigms. Denying teleology in nature is as much a matter of imposing our concepts on something beyond us as attributing teleology to nature is, so the charge of anthropomorphism seems rather irrelevant.

To answer that one has to account for phenomenal experience, which is a complex and speculative subject; but Damasio's neurologically-based account of the how the (evolutionarily) early mappings of the internal milieu to a dynamic body image in the brain, gave rise to primordial feelings ('Self Comes to Mind' Ch.4 'The Body in Mind', p.89), which subsequent evolution elaborated in richness and sophistication, and yet the primitives of pain and pleasure continued to be as powerful and dominating as ever, suggest that they are extremely effective and successful behavioural drivers.

That actually looks very interesting, thank you. From the Google preview, I can see that I'm going to find it both intriguing and frustrating--and for the same reason. He seems like a bit of a phenomenologically inspired neuroscientist, but we continentalists are nothing if not analytically imprecise!

I don't have a problem with an evolutionary account of consciousness. The problem is the logical jump from a physicalist ontology to even primordial feelings. And if he's going to attribute an unconscious will to live to cells to make sense of it, then we're back to Aristotle again. (And on the other hand, if it's anthropomorphic language, then "primordial feelings" might as well be also, and we're left wondering when nothing becomes something.)

By abstraction, I typically mean a generalisation or indirection of properties or relationships (my usage is probably influenced by my career in object-based software design, where abstract types are ideals or templates that cannot be directly instantiated, and to which concrete instantiable types have an 'is-a' relationship; e.g. an apple is a fruit). Most people don't seem to have difficulty understanding what I mean, but if you have trouble following, just ask and I'll try to explain.

I understand you just fine. The problem is that dialectical clarity is a prerequisite for any fruitful discussion of metaphysics. Otherwise you can end up slipping around like a Marxist.

I'm concerned with referring to consciousness as an abstraction if by abstraction, we mean generalization. There is nothing general about it, and whatever processes are underlying it, it's not indirect.

I don't know, nor, it seems, does anyone else; but they do.

We technically don't know that, since if there is no such thing as the purely physical, then the purely physical is obviously not giving rise to a self-sense at all. This is specifically a problem for physicalism, and not one that other ontologies share. If dual-aspect monism is correct and mental and physical states are two perspectives on the same thing, then there's really no deep paradox here at all.

As for a fundamental divide between objective and subjective, there are traditions out there that deny that altogether. (Vedantic non-dualism in particular is famous for this, though all religious traditions tend to do this to one degree or another, placing transcendent Being beyond objectivity and subjectivity. I think they're correct to do so, human concepts being what they are.)

 

[Bobber]

I don't think computers/robotics can ever be self-aware. Any way you want to cut it everything within them is programing. Some have suggested robots should some day be given the right to vote. How nice! Who does the programing? Guess they get more votes than their one. How could robotics ever understand morals except the one who programed it? It has to resort to it's "A" is good, "B" is bad and take actions based on what's put into it.

 

[Strathos]

I don't think computers/robotics can ever be self-aware. Any way you want to cut it everything within them is programing. Some have suggested robots should some day be given the right to vote. How nice! Who does the programing? Guess they get more votes than their one. How could robotics ever understand morals except the one who programed it? It has to resort to it's "A" is good, "B" is bad and take actions based on what's put into it.

This is a very simplistic reductionist understanding of cognition. All of the activity in a human brain can also be reduced to individual sequences of firing synapses, which could theoretically be run on a Turing machine with the proper instructions (programming).

 

[FrumiousBandersnatch]

Wait a moment, are you a transcendental idealist who just prefers the language of materialism as some sort of constructed subjective reality?

Lol, no. I don't really identify or align with any particular philosophical school - not a fan of such quasi-religious tribalism. I try to go with what the evidence suggests to me, informed by various philosophical ideas. If you want labels, I'm probably more naturalist than anything else.

Every attempt to superimpose human concepts upon reality is going to be anthropomorphism. This is as true for those who view nature through the language of machinery (or in modern computational terms) as it is for older paradigms. Denying teleology in nature is as much a matter of imposing our concepts on something beyond us as attributing teleology to nature is, so the charge of anthropomorphism seems rather irrelevant.

All our concepts are human concepts; what I mean by anthropomorphism is applying the high-level concepts we use for human behaviours to non-human entities/phenomena; the attribution of human-like thought processes to entities/phenomena that lack the processing complexity to support it. It's fine as poetic metaphor, but a mistake to take literally, IMO.

But broadly, I agree with you - which is why I think teleological descriptions should be reserved for agents that can plan their actions.

I don't have a problem with an evolutionary account of consciousness. The problem is the logical jump from a physicalist ontology to even primordial feelings. And if he's going to attribute an unconscious will to live to cells to make sense of it, then we're back to Aristotle again. (And on the other hand, if it's anthropomorphic language, then "primordial feelings" might as well be also, and we're left wondering when nothing becomes something.)

He doesn't get that specific, and certainly doesn't invoke the 'will to live'.

I'm concerned with referring to consciousness as an abstraction if by abstraction, we mean generalization. There is nothing general about it, and whatever processes are underlying it, it's not indirect.

I think consciousness is an abstraction in as much as it's not well-defined and is used to describe a variety of related mental activities.

As for a fundamental divide between objective and subjective, there are traditions out there that deny that altogether. (Vedantic non-dualism in particular is famous for this, though all religious traditions tend to do this to one degree or another, placing transcendent Being beyond objectivity and subjectivity. I think they're correct to do so, human concepts being what they are.)

I know people use such language, but I don't know what it means; I don't know what 'beyond theist and atheist' means, or 'beyond causal and random'...

Seems to me that one can assemble words in all kinds of evocative or poetic ways, but they don't necessarily have meaning.

Maybe you can throw some light on what 'beyond objectivity and subjectivity' is supposed to mean?

 

[durangodawood]

...I think teleological descriptions should be reserved for agents that can plan their actions....

I thought this was built into the idea of telos: the capacity to conceptualize a future.

The whole idea seems meaningless without that.

 

[FrumiousBandersnatch]

I thought this was built into the idea of telos: the capacity to conceptualize a future.

The whole idea seems meaningless without that.

Yes; but not everyone agrees - as Silmarien said, some, e.g. Neo-Aristotelians, invoke immanent teleology as a principle of nature, and deny that it relates to agency.

I can understand people interpreting everything in terms of the agency of deistic design, or pantheistic agency, but as you say, purpose without direct or indirect agency seems meaningless...

 

[durangodawood]

Teleology is associated with agency amongst opponents of teleology, and of course amongst the general public. If you actually look at the Neo-Aristotelians who invoke immanent teleology as a principle of nature, the first thing they do is deny that teleology has anything to do with agency at all.

Then whats left to distinguish between telos and just what happens as natural forces play out over time?

telos seems to me like a type of desire, which requires a desiring subject.

 

[Silmarien]

Lol, no. I don't really identify or align with any particular philosophical school - not a fan of such quasi-religious tribalism. I try to go with what the evidence suggests to me, informed by various philosophical ideas. If you want labels, I'm probably more naturalist than anything else.

I don't consider it tribalism to aim for a cohesive philosophical system that does not contradict itself wildly. Once you start invoking Kant and saying that all we have access to are our own concepts, I don't know how you can be a naturalist. Naturalism is as much a case of imposing our limited human ideas on an unknowable external world as any other system is, if not more so.

All our concepts are human concepts; what I mean by anthropomorphism is applying the high-level concepts we use for human behaviours to non-human entities/phenomena; the attribution of human-like thought processes to entities/phenomena that lack the processing complexity to support it. It's fine as poetic metaphor, but a mistake to take literally, IMO.

But broadly, I agree with you - which is why I think teleological descriptions should be reserved for agents that can plan their actions.

Immanent teleology doesn't involve attributing human-like thought processes to anything that is non-human. That is a caricature, not the actual position. Teleology means that all substances are directed towards ends by virtue of their intrinsic nature. Essentialists would attribute active tendencies towards outcomes to physical substances, instead of seeing natural phenomena as passively following laws imposed upon nature.

I would reserve actual agency to agents, but immanent teleology is something different, and to deny it and insist upon a mechanistic view of nature is really reenvisioning nature through the lens of a specific era in human history. Every clock metaphor is a form of anthropomorphism.

He doesn't get that specific, and certainly doesn't invoke the 'will to live'.

Perhaps he was speaking allegorically (I've only read the Google preview), but the word "will" definitely gets thrown around at the cellular level. The language in the book is very evocative in general, and I'm unsure how to interpret that. It looks like an interesting book, but I don't think it's going to address the specific issues I have in a particularly useful manner. (I don't think he means to, though, so that's okay.)

I know people use such language, but I don't know what it means; I don't know what 'beyond theist and atheist' means, or 'beyond causal and random'...

Seems to me that one can assemble words in all kinds of evocative or poetic ways, but they don't necessarily have meaning.

Maybe you can throw some light on what 'beyond objectivity and subjectivity' is supposed to mean?

Well, objectivity and subjectivity are concepts that are meaningless except in conjunction. If there is no subject, then there is no object either--in the absence of subjective experience, we don't have objective reality. All we have is reality.

It's the sharp divide between objectivity and subjectivity that really troubles me, since it is very dualistic and I don't see how you can keep the two things separate without ending up at a sort of substance dualism where reality has two sides to it--viewed from within and viewed from without. There is of course debate on how to solve this problem, since there are some materialists who want to eliminate the subjective, whereas idealists would instead deny the objective. Either way, if the divide breaks down, so do both concepts.

"Beyond atheism and theism" is much easier to explain, because every non-theistic concept of the divine fits in this category. The quintessential one would, again, be Advaita Vedanta, with its elimination of the world until all that exists is Brahman. Monistic ontologies are interesting like this, since when reality is reduced to one substance--call it God, call it Being, call it matter, call it whatever you wish--things tend to just blend together.

I thought this was built into the idea of telos: the capacity to conceptualize a future.

The whole idea seems meaningless without that.

No, τέλος just means "end," as in purpose or goal. It's not about conceptualization, just about goal-directedness. We've conflated it with τέχνη (techne), which is intentional craftsmanship, but the two concepts are distinct. If the natural order is a matter of substances acting according to their natures instead of according to forces and laws that have been imposed upon them, then we're getting into the realm of teleology in the Aristotelian sense. (It's actually one of the most naturalized metaphysics, but naturalists tend not to like it.)

 

[durangodawood]

....No, τέλος just means "end," as in purpose or goal. It's not about conceptualization, just about goal-directedness. We've conflated it with τέχνη (techne), which is intentional craftsmanship, but the two concepts are distinct. If the natural order is a matter of substances acting according to their natures instead of according to forces and laws that have been imposed upon them, then we're getting into the realm of teleology in the Aristotelian sense. (It's actually one of the most naturalized metaphysics, but naturalists tend not to like it.)

Is there any natural process that does not have a telos?

 

[Silmarien]

Is there any natural process that does not have a telos?

No, because telos is built into the idea of a process in the first place. If the process has no telos, it has no outcome. It isn't a process anymore.

 

[durangodawood]

No, because telos is built into the idea of a process in the first place. If the process has no telos, it has no outcome. It isn't a process anymore.

If telos is just 'what-happens', then why invoke it at all except to spray the intellectual room with the odor of meaning?

 

[Silmarien]

If telos is just 'what-happens', then why invoke it at all except to spray the intellectual room with the odor of meaning?

It isn't even specifically about meaning--it's about the correct framework for interpreting what we see occurring in the natural work. If you want a practical reason, metaphysical assumptions do play a role in how we conduct science and the sorts of possibilities we allow for. Einstein's determinism, for example, made him pretty hostile to the very theories he was developing. So if an Aristotelian conception of nature is a better, more flexible framework for viewing reality than more modern ones, then people who subscribe to it can be expected to see possibilities that other people miss.

There are other issues too, though, like the concept of biological value. If teleology and ends-directed behavior is a principle of nature and not something that we are imposing upon nature, then the idea that things can be good and bad in terms of how they relate to a certain goal makes sense again, and you can get to a coherent theory of moral realism. (This is why I go after pleasure and pain, because once phenomenal sensations that can be assigned genuine value are supervening on physical processes whose value is merely an illusory byproduct of natural selection, materialism has a great deal to explain. I'm unconvinced it can without tossing in flowery language and insisting it's metaphorical or changing the subject and hoping nobody noticed. Or claiming that everything is illusory all the way down, but then you've eliminated rationality, which certainly has consequences of its own.)

 

[durangodawood]

It isn't even specifically about meaning--it's about the correct framework for interpreting what we see occurring in the natural work. If you want a practical reason, metaphysical assumptions do play a role in how we conduct science and the sorts of possibilities we allow for. Einstein's determinism, for example, made him pretty hostile to the very theories he was developing. So if an Aristotelian conception of nature is a better, more flexible framework for viewing reality than more modern ones, then people who subscribe to it can be expected to see possibilities that other people miss.

There are other issues too, though, like the concept of biological value. If teleology and ends-directed behavior is a principle of nature and not something that we are imposing upon nature, then the idea that things can be good and bad in terms of how they relate to a certain goal makes sense again, and you can get to a coherent theory of moral realism. (This is why I go after pleasure and pain, because once phenomenal sensations that can be assigned genuine value are supervening on physical processes whose value is merely an illusory byproduct of natural selection, materialism has a great deal to explain. I'm unconvinced it can without tossing in flowery language and insisting it's metaphorical or changing the subject and hoping nobody noticed. Or claiming that everything is illusory all the way down, but then you've eliminated rationality, which certainly has consequences of its own.)

What I'm getting here is that telos is useful in 2 ways: As a mind-opening proposition. And as useful in forming a preferred sort of moral philosophy.

But these sort of by-products of holding to telos dont get me to the reality of telos (at any lever lower than for time-conscious beings).

Seems to me there's a vastly meaningful difference between 'how things behave' and 'what this being intents for this action'.

 

[FrumiousBandersnatch]

I don't consider it tribalism to aim for a cohesive philosophical system that does not contradict itself wildly.

I agree.

Once you start invoking Kant and saying that all we have access to are our own concepts, I don't know how you can be a naturalist.

I don't follow - as I understand it, if I had invoked Kant, he still allows for intuitions that are non-conceptual; so, taking an alternate meaning, what concepts, other than our own, do we have access to? and why would only having access to our own concepts disallow naturalism?

Naturalism is as much a case of imposing our limited human ideas on an unknowable external world as any other system is, if not more so.

Yes, of course. All descriptions do that.

Immanent teleology doesn't involve attributing human-like thought processes to anything that is non-human. That is a caricature, not the actual position. Teleology means that all substances are directed towards ends by virtue of their intrinsic nature. Essentialists would attribute active tendencies towards outcomes to physical substances, instead of seeing natural phenomena as passively following laws imposed upon nature.

I don't have a problem with 'active tendencies' - electrostatic attraction and repulsion are active tendencies, gravity is an active tendency; I just don't see that invoking purpose has any useful meaning for such things.

Physical laws are inferential and descriptive, not prescriptive.

It's the sharp divide between objectivity and subjectivity that really troubles me, since it is very dualistic and I don't see how you can keep the two things separate without ending up at a sort of substance dualism where reality has two sides to it--viewed from within and viewed from without.

That's how I 'see' it (), but I wouldn't call it substance dualism. It's the same substance, but there's a difference between the experience of being that substance and the experience of observing it from an external viewpoint.

 

[Silmarien]

What I'm getting here is that telos is useful in 2 ways: As a mind-opening proposition. And as useful in forming a preferred sort of moral philosophy.

But these sort of by-products of holding to telos dont get me to the reality of telos (at any lever lower than for time-conscious beings).

Seems to me there's a vastly meaningful difference between 'how things behave' and 'what this being intents for this action'.

I think you're looking at it from more of a pragmatic position: what makes for a useful (but ultimately fictional) way of describing the world, for some sort of additional purpose. But Neo-Aristotelians don't say that telos is a useful concept. They say that it's a principle of nature.

What they're doing is challenging the picture of nature that has been dominant in the West for the past 500 years. One of my favorite claims coming out of this school of thought is that the very idea of a law of nature cannot be separated from a deistic conception of reality, and that an atheistic ontology actually requires viewing material substances as having intrinsic natures.

I don't follow - as I understand it, if I had invoked Kant, he still allows for intuitions that are non-conceptual; so, taking an alternate meaning, what concepts, other than our own, do we have access to? and why would only having access to our own concepts disallow naturalism?

Depends on what you mean by naturalism. If you're just talking about methodological naturalism and sticking to scientific investigation as a way of improving our concepts, while being silent on what reality is independent of human concepts, then you're fine.

If you're moving into the realm of metaphysical naturalism and saying that reality does behave in a specific manner, then you're moving beyond the limitations placed upon human knowledge by our own concepts. Because what is naturalism except the claim that a complete scientific Theory of Everything would provide a full description of reality? The Kantian really ought to say no, because a Theory of Everything is still in the realm of the conceptual, and cannot be expected to capture reality.

You seem to jump between the two views in a way that is hard to follow.

 

[FrumiousBandersnatch]

I think you're looking at it from more of a pragmatic position: what makes for a useful (but ultimately fictional) way of describing the world, for some sort of additional purpose. But Neo-Aristotelians don't say that telos is a useful concept. They say that it's a principle of nature.

What they're doing is challenging the picture of nature that has been dominant in the West for the past 500 years. One of my favorite claims coming out of this school of thought is that the very idea of a law of nature cannot be separated from a deistic conception of reality, and that an atheistic ontology actually requires viewing material substances as having intrinsic natures.

They can claim anything they like, but the onus is on them to show why I should take their claim seriously - unless it's just poetic contrarianism.

Depends on what you mean by naturalism. If you're just talking about methodological naturalism and sticking to scientific investigation as a way of improving our concepts, while being silent on what reality is independent of human concepts, then you're fine.

If you're moving into the realm of metaphysical naturalism and saying that reality does behave in a specific manner, then you're moving beyond the limitations placed upon human knowledge by our own concepts. Because what is naturalism except the claim that a complete scientific Theory of Everything would provide a full description of reality? The Kantian really ought to say no, because a Theory of Everything is still in the realm of the conceptual, and cannot be expected to capture reality.

You seem to jump between the two views in a way that is hard to follow.

Like I said, I don't identify with any particular philosophical position; however, I don't see a rational alternative to methodological naturalism for learning about the world, and I see no plausible evidence of the supernatural, but plenty of evidence suggesting that claims of the supernatural have, or can have, natural explanations, so a naturalistic ontogeny suggests itself. But equally, I don't think it's reasonable to make claims of certainty, and though we can posit 'ultimate' explanations like the radical Platonism of Tegmark's 'Mathematical Universe', they're probably beyond the reach of methodological naturalism.

In the absence of evidence, parsimony seems a reasonable precautionary principle to apply. To paraphrase Feynman, "We should not be afraid to admit we don't know, and it's better to have questions that can't be answered than answers that can't be questioned".