• fdrake
    6.6k


    What else do you expect if you take the attitude that we are free to construct metrics which are valid in terms of our own particular interests?

    This is a lot of flip-flopping Apo. At the start of the thread you advised me to read various entropy measures for context. Especially ascendancy. You must believe there's some good stuff in the entropic measures, that they provide a link between the 'generalised' entropy that you want to speak about to particular contexts. Now when I point out that their aggregation would have poor construct validity, they're no longer serving as instantiations of 'generalised' entropy and caring about their behaviour is too concerned with irrelevant detail.

    You also wrote a description of entropy and work in ATP production and digestion, you do care about instantiations to contexts. But apparently you don't care about the specifics insofar as they resist easy unification into a single sense of entropy. Salthe is quite the opposite, he goes from the specific to the general then back again. Entropies are contextualised, and their parametrizations are given some respect:

    A point of confusion between H and S should be dealt with here. That is that one must keep in mind the scalar level where the H is imputed to be located. So, if we refer to the biological diversity of an ecosystems as entropic, with an H of a certain value, we are not dealing with the fact that the units used to calculate the entropy— populations of organisms—are themselves quite negentropic when viewed from a different level and/or perspective. More classically, the molecules of a gas dispersing as described by Boltzmann are themselves quite orderly, and that order is not supposed to break down as the collection moves toward equilibrium (of course, this might occur as well in some kinds of more interesting systems). — Salthe

    moreover, the distinctions between different systems are interpreted in terms of differences between their individual energetic behaviour. Exceptions are noted. Salthe's method is, essentially, attempting to form a conceptual space whereby entropy can be discussed in general by seeing what is general within the specific contexts entropy arises. He then takes the generalisations and applies them to specific contexts to re-present and contrast the nascent sense of entropy in general with the specifics. This is achieved by the description of entropy in a highly abstract register which is constituted from a web of analogies, citations and examples (but little mathematics, which is unfortunate).

    If this sounds familiar to anyone, essentially Salthe is a phenomenologist of entropy. I have no problem with this. What I do have a problem with is taking the nascent sense of generalised entropy and treating all phenomena, especially systems, as addenda to the nascent sense. In Heideggerian terms, this is an error in thematisation: Salthe gets to speak like this since he thematises entropy and writes about entropy. You don't get to speak like this about stuff in general unless you subordinate all inquiry to the inquiry of entropy. The disclosive character of Salthe's discourse has the thematisation of entropy as a necessary constituent. He is an essayist on entropy.

    You turn metaphysically ladened discussions (read: almost all discussions) into discussions about your metaphysical system. The space of problems you engage with is not thematised in terms of the problems - apparently intentionally on your part. This is a major methodological error, it occludes the nature of the problem and questions related to it in a fog of analogised and half formed notions. It is a method subsisting in obfuscation of the original character of problems, constituted by analogic metaphorical language disclosing the wrong problematic as a theme.

    Recall this is a supposed discussion of ecosystems and people/societies as ecosystems. Instantiations of entropic measures for ecosystems are apparently 'too narrow', despite being part of the topic when viewed from an information-systems perspective. Also note that degrees of freedom means something different for both measures, and neither one can be said to have boundary conditions in the literal sense or constraints in anything like the sense that we agreed upon in terms of state spaces.

    I agree we can do just that. We can describe the world in terms that pick up on some characteristic of interest. I just say that is not a deep approach. What we really want is to discover the patterns by which nature organises itself. And to do that, we need some notion about what nature actually desires.

    It is not a deep approach to attempting to generalise entropy. But your attempts to generalise entropy could not be called a deep approach to anything but entropy. I think you, perhaps like Salthe though my verdict is still out on his sins, should attempt non-metaphorical procedural descriptions of processes more often. You did a good thing by talking about entropy and work in ATP metabolism - I would commend it more if it was more relevant to entropy and information in ecosystems. You generally characterise that as a wee system 'giving freedom' to a big one. The ecosystem measures are about the big one. Salthe has a means of problematising this kind of misapplication or mislocation - a hierarchy of problem classes relevant to systems. You should be dealing with the 'given freedom' rather than the 'freedom giving' from below.

    So you are down among the weeds. I'm talking about the big picture. Again, it is fine if your own interests are narrow. But I've made different choices.

    You are a holist, and have holist concerns. Don't be hating on the weeds for being narrow, be hating on the picture for painting weeds as insufficiently general. They implicate systems of all orders of abstractions, and you seem to have forgotten this commitment.
  • apokrisis
    7.3k
    I’ve been talking about the thermodynamic constraints that shape dissipative systems like ecologies. You have failed to show why I should care about a species diversity index. The fact that the two don’t relate nicely is only to be expected.

    Your notion of generalising entropy talk is epistemic. My interests are ontic. But best of luck with your future endeavours.
  • fdrake
    6.6k


    I was typing something in anticipation to this before you responded, funnily enough. I have ontic concerns too. This is why I'm looking to add things to my reductionist utility belt.

    You would do well by paying more attention to your procedural descriptions - looking at how entropy transfers between 'levels' of organisation. I'll give an example:

    (1) Describing the dynamical system of ATP metabolism in terms of inputs and outputs.
    (2) Defining an entropic measure, or necessary features of an entropic measure, relevant to this metabolism.
    (3) Giving a relational mechanism (not just a name of it) between this metabolism and the energetic dynamics ('metabolism') of the system which uses ATP metabolism as a constituent structure. This could perhaps be achieved through averaging energy consumptions of cells and what energy sources give the fuel for the ATP metabolism - a link between biomass and derived energies.
    (4) Give a relationship between the entropic measure in (2) and the energetic measure in (3). Like the function of analogising absence of entropy with exergy.
    (5) Looking at the distribution and behaviour of derived energies on the 'top level' of analysis, also including an instantiation of the relational mechanism (maybe something like making the energy flow to one organism or its population proportional to the conversion of biomass to ATP).
    (6) Derive an entropy measure, or characterise one, relevant to (5). Relate it to the one in (4) in such a way that the product of (3) can be parametrised in accordance with it.

    That you don't do this can be related to the error of thematisation detailed above. The specific dynamics of the 'given freedom' - the behaviour of flows within the ecosystem - is glossed over, there aren't any procedural specifications from the metabolism of ATP to the 'metabolism' of flows in an ecosystem. And what is glossed over would be an excellent contribution to the account of the 'modelling relation', as you call it, between the two and would be a very productive idea in terms of generalised entropy. This is the kind of work required in generalisation, not the automatic conversion-by-handwaving from entropy in one dynamical system to another of different parametrisation and scope. I think you usually handwave this by calling it 'coupling', too (forgetting that coupled systems have shared parameter spaces). Lo and behold, when you put a bit of work in, you can see literal coupling when your figurative sense of coupling was implicated and it looks like there's some way to take the intersection of parameter spaces such that each individual system's has a non-empty intersection with enough of the rest to make a connected network of flows.

    How is this for a compromise: you keep using your terms the way you do but put more effort into relations and procedural dynamics between different ontic energy/entropy regimes and how they describe systems.

    Productive dynamical inquiry = problematising change in relations and relations of change.
  • fdrake
    6.6k


    So "disorder" is just an application of the principle of indifference. A messy system is one in which the details don't matter. Gone to equilibrium, the system will be generic or typical.

    Except no. The principle of indifference is an idea of equidistribution over a finite set of states. A die is the model case of the principle of indifference. Roll a 1 - same probability as a 6. Even with a charitable interpretation of this, it destroys any idea of there being degrees of disorder. Even Boltzmann entropy, which explicitly uses equiprobability of microstates within a macrostate still constrains the application to a specific macrostate, not a generalised definition.

    The principle of indifference cannot be extended as equiprobability to countable or continuous state spaces - this is because a uniform distribution cannot exist on infinite sets of outcomes. To see this, imagine a sphere. Imagine a particle trapped in it, wandering about in a random walk. The probability that it is occupies a given volume within the sphere is proportional to that volume - the proportion of the set volume purported to contain the particle to the sphere's total volume. If you want an equal distribution over all possible points in the sphere and think of it this way - you end up with something that isn't even a probability density (roughly, all points are 0 probability or their probabilities sum to infinity). Specific configurations having 0 probability is consistent with this volumetric idea, however.

    Regardless ignoring the elisions and providing lots of charity, this constrains entropy to maximised entropy and destroys relative degrees of disorder and order. It also doesn't work for entropy in continuous processes - only continuous processes which have mappings to discrete outcome spaces. EG, think of a partition of the sphere above for an example, each subset with a probability proportional to its size. Their sizes sum to the volume since it's a partition - but it's no longer a probability distribution indexed to the sphere, it's indirectly related through the partition. The limiting process as set volume tends to 0 doesn't produce a density.

    I doubt you would want this in 'modelling' applications, as it's well known that uniform distributions aren't necessarily entropy maximisers. Neither are they necessarily heavy-tailed distributions, which you usually imply as playing a pivotal role in the emergence of entropy over stratified hierarchical systems. They, not surprisingly, depend on the parametrisation and possibly a partitioning procedure of the system in question.
  • fdrake
    6.6k


    It is pretty clear that I’m talking about dissipative structure. And so is ascendency. So is Salthe, Kay and Schneider, Bejan and the many other cites I offered. But your critique amounts to me failing to relate dissipative structure theory to some mundane measure of species diversity.

    Ulanowicz' ascendency can be applied to any ecosystem network parametrised in terms of flow exchange, it need not be applied to a dissipative network. To see this, you can do two things: 1, look at the formula and see there is no temporal component and 2 - do some reasoning about the behaviour of the ascendency. For a definition of ascendency for reference - it is the average mutual information of flows contributing to and coming from a node scaled by the total ecosystem throughput.

    It isn't actually so clear cut that ascendency is cashed out in terms of dissipative systems. This would occur if the ecosystem network had an outflow, so that the total throughput decreased over time. Or alternatively, if the ecosystem had an inflow so the total throughput increased over time. Whether ascendency corresponds to or can detect the dissipation of energy of an ecosystem reliably or its inflow depends not just on the total throughput - which is increasing or decreasing over time in dissipative systems - but upon how changes in total throughput manifest (or don't manifest) in changes of flow concentration or equalisation.**

    The argument that ascendency has a natural interpretation in terms of dissipation doesn't just deal with the definition of ascendency - Ulanowicz doesn't actually describe it as a measure of a dissipative structure - he describes it as a measure of 'growth of an autocatalytic system'. He has intuitions that 'more developed' ecosystems will have a higher ascendency. He explicitly constructed it as an atemporal index to allow the comparison of ecosystems over time or other indices. So the behaviour of ascendency over time (or some other index) is the thing which may or may not reflect whatever assumptions you have about dissipative systems more generally.

    That it generalises to other indices which represent gradients in other quantities is put forward in the paper he defines it in:

    The process of eutrophication for example is characterised by a rise in ascendency which is due to an overt increase in the activity of the system which more than compensates for its concomitant decrease in in its developmental status. — Ulanowicz

    So what Ulanowicz said - the prediction of rising ascendency on an increasing eutrophication gradient - is empirically falsifiable since it depicts a trend in the fuzzy concept of 'ecosystem development' using a trend in an entropic measure.

    Without lingering too long on that fact that that isn't actually correct, there's an observed upside down U shape in ascendency (increase then decrease) over an eutrophication gradient, though since the paper detailing that doesn't do an error analysis it's still up for debate - he has to at least engage with the relative strengths of the terms in the formula. He does.

    And apparently none of my criticisms are relevant since my concerns are merely 'epistemic' or 'reductive'. Really? When it turns out that your equivocations aren't sound and you gloss over too much it's fine because I don't care about the 'ontic' status of ecosystems? I think I care about that quite a lot. I've actually highlighted different notions of complexity relevant to ecosystems, presented and analysed the majority of the terms you've used in your posts. Furthermore, I've spelled out the implications in terms of understanding ecosystems that your equivocations imply; then how this hand waving dis-enables productive inquiry regarding dynamical systems and informational relations. And when possible, I've looked at the empirical results relating to the infodynamics of ecosystems. Doesn't that sound like sound ontic inquiry to you?

    In contrast: you usually forgo procedural descriptions which are the glue that bind the interrelation of composite systems. You should be detailing the relationships between different subsystems in your posts procedurally, not metaphorically. You don't do this but pretend to, you cite formulas and a wealth of background literature without ever cashing out what you say in their terms. You sample bites from them and imply everything will go your way. Even now, because the single hole in my responses to you is that I haven't explicitly engaged any dissipative systems material, you focus on that as if everything I've said wasn't relevant for another reason. So:

    we have to take on trust that you are being faithful in using and interpreting the measures. We have to take it on trust that your descriptions elucidate the relational character of different systems in a manner consistent with your references. And we have to take it on trust that your interpretations of your references are well informed and valid.

    I don't think that trust is vindicated any more.
  • fdrake
    6.6k
    A last thing, looking at ascendency as a measure of a dissipative system.

    Ulanowicz' use of 'autocatalysis' can signify a growth or amplification of some aspect of the system. This, analogically, relates to dissipative systems. It doesn't necessarily mean the measure will have the 'ampliative' property in terms of necessitating a growing measure over time (as was seen empirically). Nor will the measure necessarily decrease over time (again, empirically, that U-shape is weird). The imaginative background of the measure suggests that it will behave in that manner - however it doesn't.

    There might be an avenue of studying the system as dissipative in terms of how a unit of energy is distributed around the network. It would be possible to define a probability distribution on each node for the probability that a given quantum of energy leaves that node and goes to another, and then we add another node for 'wasted energy' which is connected as a sink to all other nodes. This would probably introduce a directed Markovian graph in some sense dual to the system. The iterations of this Markovian graph may have a steady state (equilibrium distribution), and the distance from the steady state from a particular graph could be studied.

    But there isn't a theory of dissipation built into the ascendency formula from the get go.
  • apokrisis
    7.3k
    I think you usually handwave this by calling it 'coupling', too (forgetting that coupled systems have shared parameter spaces). Lo and behold, when you put a bit of work in, you can see literal coupling when your figurative sense of coupling was implicated and it looks like there's some way to take the intersection of parameter spaces such that each individual system's has a non-empty intersection with enough of the rest to make a connected network of flows.fdrake

    This is an example of our different interests. You presume parameter spaces can be glued together. I'm concerned with the emergent nature of parameters themselves. You have your idea of how to make workable models. I'm interested in the metaphysics used to justify the basis of the model.

    So it just gets tiresome when your criticism amounts to the fact I'm not bothered about the details of model building for various applications. I've already said my focus is on paradigm shifts within modelling. And core is the difference between mechanical and organic, or reductionist and holist, understandings of causality.

    The principle of indifference cannot be extended as equiprobability to countable or continuous state spaces - this is because a uniform distribution cannot exist on infinite sets of outcomes.fdrake

    Another inventive way to miss the point I was arguing. Sure Boltzmann had a problem if the world wasn't actually atomistic and entropy was a continuous substance, a caloric.

    Without lingering too long on that fact that that isn't actually correct, there's an observed upside down U shape in ascendency (increase then decrease) over an eutrophication gradient, though since the paper detailing that doesn't do an error analysis it's still up for debate - he has to at least engage with the relative strengths of the terms in the formula. He does.fdrake

    You can have your doubts about the robustness of his approach. But again, I was responding to the OP in terms of what ecologists actually say about ecologies based on the kind metaphysics they've actually developed.

    And both Salthe and Ulanowicz argue for a three stage lifecycle - that inverted goldilocks U-curve. Regardless of how well it may or may not be cashed out in real world models, the general metaphysical level argument seems sound enough, and obvious enough, to me. If you want to critique that, then great.

    Ulanowicz describes the motivating metaphysics in: The dual nature of ecosystem dynamics, 2009 - http://izt.ciens.ucv.ve/ecologia/Archivos/ECO_POB%202009/ECOPO7_2009/Ulanowicz%202009.pdf

    The yin and yang of ecology

    By now the reader may have noticed that two countervailing tendencies are at play in the development of any dissipative structure. In one direction a continuous stream of perturbations works to erode any existing structure and coherence. Meanwhile, this drift is opposed by the workings of autocatalytic configurations, which drive growth and development and provide repair to the system.

    This tension has been noted since Antiquity. Diogenes related that Heraclitus saw the world as a continuous tearing down and building up. With the Enlightenment, however, science opted for a more Platonic view of nature as monistic equilibrium.

    Outside of science, Hegel retained Heraclitus’ view of the fundamental tension, but with significant amendment. He noted that, although the two tendencies may be antagonistic at the level of observation, they may become mutually obligatory at the next higher level. Hegel’s view is resonant with the picture of ecosystem dynamics portrayed here.

    Indeed, the second law does dissipate what autocatalysis has built up, but it has been noted that singular chance is also necessary if systems are truly to evolve over time and develop novel emergent characteristics. Looking in the other direction, complex, evolved systems can be sustained only through copious dissipation.

    The problem with this agonistic view of the natural world is that, unlike the mechanistic (Platonic) convention, dialectic like dynamics cannot be adequately represented as algorithms.

    To repeat again, mechanistic simulation models are inadequate to the task of describing ecosystems over the longer run, because the selfsame selection exhibited by autocatalysis can unpredictably replace not only components, but their accompanying mechanisms as well. Not only does the notion of mechanism defy logic, it seems also to poorly match the dynamics that actually are at play.

    So complexity is irreducible once we start talking about self-parameterising systems. Like a dissipative structure. All your fussing about a lack of particularisation of parameter spaces by me makes no sense as we are essentially - when doing science of such systems - talking about modelling as an art. We can make better or worse choices. And any choice must be guided by some grounding, if fuzzy or vague, intuition. (Such as that entropy/information/exergy/degrees of freedom/whatever are "this kind of generic thing or process".)

    You are continually jumping to the position of there being a right way to measure the world. But it is basic to a particular group of theoretical biologists I respect - Salthe, Pattee, Rosen, Ulanowicz - that measurement is an informal act. An art or exercise of good judgement. And this is because the world of interest is inherently non-linear - it has a complexity that is irreducible.

    Ulanowicz then goes on to talk about the lifecycle model which is relevant to the OP. He makes the point that his ascendancy has this dualistic dynamic. The trade-off is between the organisational power of a system - its useful order - versus the systems overhead needed to physically instantiate that pattern of organisation.

    The chief advantage of using information theory to describe organization is that it allows one also to quantify the opposite (or complement) to information in similar fashion. Whence everything
    that is disordered, incoherent and dissipative in the same network can be captured by a related, non-negative variable called the system’s overhead...Furthermore, a system’s ascendency and overhead sum to yield its overall capacity for development.

    The actual pattern of order is the result of two opposing tendencies: In an inchoate system (one with low a), there are manifold opportunities for autocatalytic cycles to form, and those that arise create internal constraints that increase A (and thereby abet a). This tendency for a to grow via autocatalysis exists at all values of a. The role of overhead however, changes as the system progresses toward higher
    a.

    In inchoate systems (low a), it is ˚ that provides the opportunities for new cycles to form. In doing so it abets the tendency to increase autocatalysis. However, in systems that are already highly developed
    (a ≈ 1), the dominant effect of ˚ becomes the disruption of established feedback loops, resulting in a sudden loss of organized performance. (The system resets to much a lower a.)

    So at high a, ˚ strongly opposes further increase in a. Presumably, a critical balance between the countervailing roles of ˚ exists near the value of a at which the qualitative role of ˚ reverses.

    Or as the Wiki page sums it up more simply...

    Originally, it was thought that ecosystems increase uniformly in ascendency as they developed, but subsequent empirical observation has suggested that all sustainable ecosystems are confined to a narrow "window of vitality" (Ulanowicz 2002).

    Systems with relative values of ascendency plotting below the window tend to fall apart due to lack of significant internal constraints, whereas systems above the window tend to be so "brittle" that they become vulnerable to external perturbations.

    So my reply to the OP was to point out what I find to be a metaphysically reasonable account of a natural lifecycle approach to systems. It is a model developed for describing ecological systems, but both Salthe and Ulanowicz say it does extrapolate to the political and economic levels of sociological analysis.

    Thus I demonstrated that within my Peircean/organicist kit-bag of well grounded metaphysical concepts, here is a sound argument that has been already advanced.

    Then check it out and you see Ulanowicz is pretty exercised by the Hegelian logic which his approach employs. He is very concerned with the Rosen modelling relation and what it says about the informality of acts of measurement and the incommensurability of mechanistic models to irreducibly complex worlds. He employs the very same metaphysical kitset as me. (Well up to a point. Ulanowicz is a good Catholic and so we differed on the theistic slant he was working towards on the sly. :) )

    You then come along with the intent of nit-picking away, complaining that I don't follow through from a general metaphysical view to the particularity of every possible kind of entropy/information/exergy/degrees of freedom/whatever model.

    Well, like ... whatever.
  • fdrake
    6.6k


    You're getting more interesting as I press you more. Thanks for the more detailed reply.
  • charleton
    1.2k
    Succession does not always lead to more complexity. It depends on the specific case of the environment.
    For example in the post ice age landscape of the South Downs of England the tundra led to a range of scrub, bushes, heathland and trees. The species diversity increases in some instances, but can as easily become less complex by bearing fewer species.
    Ash, elm, beach, hazel, will eventually succumb to the climax vegetation which in this case is oak woodland so dense as to make many larger herbivores seek life elsewhere; all the scrubs and less long lived trees will have to give way to the oak.
    Then comes a human and destroys everything but wheat fields. The soil erodes away and any growth at all depends on the application of chemical fertilisers.
    This is not to say that humans always result in the destruction of the ecology. Aborigines of New South Wales used to practice fore-stick farming. Where the climax vegetation has led to natural monoculture, setting fire to the landscape can have a massive effect of increasing species diversity and the appearance of nut, and fruit bearing plants.
  • Metaphysician Undercover
    13.2k
    Digging in the weedsfdrake

    That's it! That's the solution to the whole problem. Get into the garden and dig the weeds. First, we have to determine what exactly is a weed.
  • apokrisis
    7.3k
    Here's a book chapter arguing the same thing in a more general philosophical way.

    https://www.researchgate.net/profile/Robert_Ulanowicz/publication/292610642_Enduring_metaphysical_impatience/links/56b00ad608ae9c1968b490b7/Enduring-metaphysical-impatience.pdf?origin=publication_detail

    I like two key points. Natural systems are irreducibly complex because they feed off their own accidents. The traditional mechanical view wants to separate the formal laws constraining systems from the material accidents composing systems. But nature includes the accidental or spontaneous in the very thing of forming its laws, or constraining regularities.

    This is beautifully Peircean. The "stuff" of the world isn't some disconnected universal machinery. It's Being includes its accidents as part of what makes any laws.

    The second point is Elsasser's combinatrics argument which says that even a computable universe is very quickly an intractably computable one. Really, the Universe is always "a one off". A propensity view of statistics - as argued by Peirce and Popper - has to be fundamental.

    Put the two together and every state of the Cosmos is an instance of one-off chance - or at least as much as it is its traditional "other" of a deterministic, law-constrained, mechanism.

    Ulanowicz sums up the dichotomy of this "ordering vs the disordering" tendency in complexity thus.....

    Elsasser argued that nature is replete with one-time events - events that happen once and never occur again. Accustomed as most investigators are to regarding chance as simplistic, Elsasser's claim sounds absurd. That chance is always simple, generic, and repeatable is, after all, the foundation of probability theory.

    Elsasser, however, used combinatorics to demonstrate the overwhelming likelihood of singular events. He reckoned that the known universe consists of somewhere on the order of 10^85 simple particles. Furthermore, that universe is about 10^25 nanoseconds in age. So at the outside, a maximum of 10^110 simple events could possibly have transpired since the Big Bang.

    Any random event with a probability of less than 1 in 10^110 of recurring simply won't happen. Its chances of happening again are not simply infinitesimally small; they are hyper-infinitesimally small. They are physically unreal.

    That is all well and good, one might respond, but where is one going to find such complex chance? Those familiar with combinatorics are aware, however, that it doesn't take an enormous number of distinguishable components before the number of combinations among them grows hyper-astronomically.

    As for Elsasser's threshold, it is reached somewhere in the neighborhood of seventy-five distinct components. Chance constellations of eighty or more distinct members will not recur in thousands of lifetimes of the universe.

    Now it happens that ecologists routinely deal with ecosystems that contain well over eighty distinct populations, each of which may consist of hundreds or thousands of identifiable individual organisms. One might say, therefore, that ecology is awash in singular events, They occur everywhere, all the time, and at all scales.

    None of which is to imply that each singular event is significant. Most simply do not affect dynamics in any measurable way; otherwise, conventional science would have been impossible. A few might impact the system negatively, forcing the system to respond in some homeostatic fashion.

    A very rare few, however, might accord with the prevailing dynamics in just such a way as to prompt the system to behave very differently. These become incorporated into the material workings of the system as part of its history. The new behavior can be said to "emerge" in a radical but wholly natural way that defies explanation under conventional assumption.

    So good luck to any science based on a mechanistic metaphysics that presumes accidents are simply uncontrolled exceptions that can be hidden behind a principle of indifference. Yet also the universe does have lawful regularity. It incorporates its accidents into the habits it then forms.

    This is just a far more interesting story of the Cosmos than the usual one that pictures it as a mathematical clockwork - the one where science is reduced to a collection of measurement protocols rather than scientific measurement being the crafty art we know it to be.
  • Streetlight
    9.1k
    Not to put too fine a point on it, but you come across as a bit of a hysteric (in the strict psychoanalytic understanding of it!), looking for the grand-plan behind it all that explains everything else - but... there really is no big Other here, I just genuinely like exploring these ideas and sketching connections where I can find them (if I can find them): this is bricoleur, not conspiracy - give me some credit for my innocence!

    But still -

    To get political: isn't not too closed, not too open, being self-regulating while allowing lines of flight - i mean isn't that, in a perfect nutshell, neoliberalism?csalisbury

    Again, I see where you're coming from - although come on, first I'm a conservative, now I'm a neoliberal? - but again, I think you're obscuring the specificity of neoliberalism, which yes, does promote a kind of 'sustained growth' narrative but - only along a single dimension: specifically, that of market values - efficacy, outcome KPIs, best-practices - being the only things that neoliberalism can 'see'/is sensitive to. I think one of the nice things about seeing the world in ecological terms is precisely the fact that it forces one to take into account so-called 'externalities' - but better, it doesn't/can't even discriminate between 'externalities' and er, not-externalities because the whole point is that everything belongs to an ecosystem.

    And look, I also acknowledged previously that one ought to not confuse tactics with strategy: perhaps, if we want to build a new, better world, it's necessary to burn this one down. Perhaps this is not the society worth trying to make better. One ought to separate what needs to happen from what one ought to happen - this perhaps being a distinction between politics and ethics.

    So all that being said, acknowledging I can't keep up with the math, I'm still confident enough to engage the OP on its own terms which are, I believe, metaphorical. Which isn't to say I think you think that self isn't literally an ecosystem - I believe you do, and I probably agree - but that I think the significance of this way of looking at the self ultimately relies on - and is motivated by- what can be drawn from it conceptually. It's about drawing on empirically-sourced models to the extent that they facilitate conceptual considerations. It's metaphorical in the literal sense that we're transporting some way of thinking from one level to another.csalisbury

    So yeah, this is a fair way to read the OP - I probably prefer it in fact. But then, I'd also say that this just is a kind of model of philosophy as such: drawing conceptual inferences, re-contextualizing the significance of empirical findings, etc, etc.
  • Streetlight
    9.1k
    Succession does not always lead to more complexity. It depends on the specific case of the environment.
    For example in the post ice age landscape of the South Downs of England the tundra led to a range of scrub, bushes, heathland and trees. The species diversity increases in some instances, but can as easily become less complex by bearing fewer species.
    Ash, elm, beach, hazel, will eventually succumb to the climax vegetation which in this case is oak woodland so dense as to make many larger herbivores seek life elsewhere; all the scrubs and less long lived trees will have to give way to the oak.
    charleton

    This is great! And my immediate thought is to relate this to the longer term dynamics of capitalism, in which certain established actors in the system co-opt resources and so deprive the 'smaller' actors of potential for growth so that they are either driven out or forced to live in what amounts to environmental ghettos. If I've been thought anything here it's that it might be interesting to map and taxonomize the differing trajectories of ecological systems and see what they might teach us about social ones.
  • Streetlight
    9.1k
    I didn't get a chance to read everything, but in the case of thermodynamic systems, the evolution of any given system is determined toward a state of equilibrium, and ergodicity attempts to ascertain the averages of behaviour within a system (transformations, arbitrary convergence, irreducibility etc) and political systems are an attempt to order the nature of Hobbesian chaos. I really like this:TimeLine

    I think it's an interesting question to see where Hobbes might fit into all of this. I mean, the first and most usual criticism of Hobbes is that his so-called state of nature doesn't look anything like nature; or at least, it only captures an incredibly narrow slice of it, which is otherwise replete with myriad other possibilities of social relation. And in this sense I think ecology is actually incredibly well positioned to show exactly why approaching the political through a Hobbesian lens is so deplorably misleading, insofar as ecosystems do in fact have all these differing and interesting trajectories which are constrained and enabled by all sorts of various parameters.

    I mean, this is the point of 'open systems', that, with enough energy flow and some constraints thrown in, you end up with order, or more specifically, self-organization. The social contract, insofar as it is imposed 'from above', is more or less an inversion of this view. Deleuze writes of how Hume, for instance, undertakes exactly such an inversion of Hobbes:

    "The fault of contractual theories is that they present us with a society whose essence is the law, that is, with a society which has no other objective than to guarantee certain preexisting natural rights and no other origin than the contract. Thus, anything positive is taken away from the social, and instead the social is saddled with negativity, limitation, and alienation. The entire Humean critique of the state of nature, natural rights, and the social contract, amounts to the suggestion that the problem must be reversed.... This understanding of the institution effectively reverses the problem: outside of the social there lies the negative, the lack, or the need. The social is profoundly creative, inventive, and positive." (Empiricism and Subjectivity); I think this is exactly the model that ought to be adopted.

    Or else there is Kropotkin's wonderful anarchist line that "accustomed as we are by hereditary prejudices and absolutely unsound education and training to see Government, legislation and magistracy everywhere around, we have come to believe that man would tear his fellow man to pieces like a wild beast the day the police took his eye off him; that chaos would come about if authority were overthrown during a revolution. And with our eyes shut we pass by thousands and thousands of human groupings which form themselves freely, without any intervention of the law, and attain results infinitely superior to those achieved under governmental tutelage." (The Conquest of Bread).
  • fdrake
    6.6k


    This is an example of our different interests. You presume parameter spaces can be glued together. I'm concerned with the emergent nature of parameters themselves. You have your idea of how to make workable models. I'm interested in the metaphysics used to justify the basis of the model.

    So it just gets tiresome when your criticism amounts to the fact I'm not bothered about the details of model building for various applications. I've already said my focus is on paradigm shifts within modelling. And core is the difference between mechanical and organic, or reductionist and holist, understandings of causality.

    I think you're painting my objections as mechanistic and reductionist because I haven't adequately communicated what my objections to your use of 'entropy in general' are. This boils down to two ideas, one is that entropy is a parametrised concept and the other is that you provide little information on how different senses of entropy are at work in your ideas of it.

    We can discard the discussion of thematisation errors, since it doesn't seem to interest you. We also have more than enough to talk about.

    My motivations in pointing out different senses of entropy aren't supposed to be aimed at demonstrating either the impossibility or futility of generating a generalised conception of entropy. Rather, they are pointing out difficulties in generalisation which I believe are relevant. Another purpose of using them as examples, as well as ascendency, is to show that there are different senses of entropy as they apply to different systems. This isn't to say there aren't general laws or a sufficient description of translation of one type of entropy to another, it's to say that there is a need to make the hows of entropy transfer between constitutive systems in a complex one a part of your accounts.

    Let's take a tangent on the idea of transduction. Transduction is the change of one type of energy to another of a qualitatively different but quantitatively related sort. Hold a ball out at arm's length and suspend it against gravity. The position the ball is kept it relative to the floor imbues it with potential energy. When the ball is released, the potential energy is converted into kinetic energy as the ball falls. The description of the dynamics in terms of energies is equivalent to the description in terms of velocities and accelerations. There's also an equation that links potential and kinetic energy for the motion of the ball.

    In an ecosystem, it's unfair to expect that there will be a complete description of its energetic and entropic dynamics. I've never had the mechanical hope that they can be specified like this. Regardless, when we talk about the transfer of energy between one functional stratum of an ecosystem to another, a description of the transduction of energy is the right kind of account. This could form part of a series of equations - which are likely to be too precise to capture all relevant stuff -, or a part of a historical description of the procedures constituting the transduction - which is likely to have an insufficient understanding of the specifics.

    This is probably easier to understand in the context of economics. To a first approximation, there are two kinds of empirical analysis in economics. One is based on the mathematical analysis of trends and betting strategies, one is based on a historical description of what happened and can utilise mathematics in a less central role. Future prospects are analysed in either the predictive distribution of a mathematical model or qualitative necessities or high probabilities of event occurence (EG, the tendency of the rate of profit to fall in Marx or Say's Law)

    James Simmons and Marx are good examples of each school.

    I'll take an example of how you describe the dynamics of entropy and complexity in a system.

    Canopy succession is an example. Once a mighty oak has grown to fill a gap, it shades out the competition. So possibilities get removed. The mighty oak then itself becomes a stable context for a host of smaller stable niches. The crumbs off its feeding table are a rain of degrees of freedom that can be spent by the fleas that live on the fleas.

    'So possibilities get removed' - how? Which possibilities are closed?
    'The crumbs off its feeding table are a rain of degrees of freedom that can be spent by the fleas that live on the fleas' - the canopy opens up, what degrees of freedom does this create? How do those degrees of freedom get turned into degrees of freedom for certain organisms? Which organisms? What properties do those organisms have? How do the 'degrees of freedom' in the 'crumbs' relate to the 'smaller stable niches', in what manner do they 'rain'? In what manner are they 'spent'?

    When I actually try and analyse your descriptions of how things work, there's so many ambiguities - this is an essential part of your posting style in which every procedure of 'entropy transduction' they contain is named but not specified. Your key terms: constraint, entropy, degrees of freedom, possibilities, information, symmetry breaking, dissipation; only obtain their sense in your posts through the background knowledge of their analogies in different contexts. When it actually comes to describing the hows - procedural system dynamics - the approach you take is holistic but empty of content.

    Over the course of this discussion, I've probably read 8 or so hours worth of papers and I still don't have a clue on how you actually think things work. When you cite things, they don't actually flesh out the procedural descriptions in your posts. You want to be giving an account of how entropy flows over systems and how it changes each component and introduces new components. The general structure of your posts in this regard is to substitute in a concept related to entropy in an unclear manner and use the web of associations between the varied concepts of entropy and their contexts to flesh out the rest.

    The purpose construct validity has in my argument is to highlight this fact. The way you use the key terms in your posts isn't cashed out by the background references. You analogise too much and specify too little.

    I like two key points. Natural systems are irreducibly complex because they feed off their own accidents. The traditional mechanical view wants to separate the formal laws constraining systems from the material accidents composing systems. But nature includes the accidental or spontaneous in the very thing of forming its laws, or constraining regularities.

    Great! Makes sense! Parametrisations and contextualised procedural descriptions are ways of studying the behaviour of ecosystems. Mathematical models of them aren't complete pictures either, they're supposed to elucidate certain behaviours in parts of ecosystem behaviour. If someone actually believed a particular mathematical model is a perfect description of all of the ecosystem dynamics that it was applied to they'd almost certainly be wrong.

    Regardless, the relationship of parameters in ecosystems and how subsystems become parametrised isn't something you can offload to the literature through your analogical web. Talking about these things is what it means to talk about the dynamics of ecosystems. You neglected to mention that one of the subsections of the paper you quoted from Ulanowicz is titled:

    The Return of Law to Ecology?

    which then goes on to analyse the dynamical properties of ascendency in a partly mathematical and partly phenomenological (in the scientific sense) manner.

    So good luck to any science based on a mechanistic metaphysics that presumes accidents are simply uncontrolled exceptions that can be hidden behind a principle of indifference. Yet also the universe does have lawful regularity. It incorporates its accidents into the habits it then forms.

    Again yes, I broadly agree with this. Ulanowicz' use of networks - not just quantities - to analyse the behaviour of ecosystems is exactly the kind of mathematical analysis that makes sense in this context. Graphs and networks are highly generalised and already have notions of 'flow','resistance' and 'variation' so long as their edges are weighted. The specific manifestations of the graphs in applied mathematical modelling of ecosystems are pretty bad at predicting their future except sometimes in the short term. So, if you want to think of ecosystems as flow networks, there's a lot of abstract generality there to exploit.

    So, we don't have any irreconcilable methodological disagreements. I don't think this is a matter of mechanism vs organicism or reductionism vs holism. The particular beef I have is that you provide poetic descriptions of systems behaviour without fleshing out the details, and this destroys the credibility of the accounts.

    So, could you please provide a procedural re-description of:

    Canopy succession is an example. Once a mighty oak has grown to fill a gap, it shades out the competition. So possibilities get removed. The mighty oak then itself becomes a stable context for a host of smaller stable niches. The crumbs off its feeding table are a rain of degrees of freedom that can be spent by the fleas that live on the fleas.

    In a manner that answers the questions I have:

    Which possibilities are closed?
    What degrees of freedom does this create? How does it create rather than destroy them? How do those degrees of freedom get turned into degrees of freedom for certain organisms? Which organisms? What properties do those recipient organisms have? How do the 'degrees of freedom' in the 'crumbs' relate to the 'smaller stable niches', in what manner do they 'rain'? In what manner are they 'spent'? How does one set of degrees of freedom in the canopy become externalised as a potential for the ecosystem by its deconstruction and then re-internalised in terms of a flow diversification?

    Being able to translate out of the abstract register to a specific example is essential to ensure the validity of your concept applications. When I tried with various entropy notions I couldn't, I still can't in terms of ascendency.
  • fdrake
    6.6k
    This is just for interest. A thing I noticed about the ecosystem flow networks is that when you analyse them in terms of flow proportions and add the nodes for the 'system source' and 'energy loss' - the energy loss node is accessible from every other node but can't be escaped. This makes the energy loss node an absorbing state. When energy is transferred around the network over time, it's very likely that this induces a dissolution of the flow network; the flows concentrate on the energy sink. This is one possible configuration where the ascendency tends to 1.

    One way of interpreting this is that the flow network in the ecosystem is metastable (figuratively). I think it's quite unlikely that there's another attractor in the flow network when there's a single absorbing state. Another is that if we looked at the evolution of the flow network over time (internal time to the network would be the application of its transition matrix to an initial state vector, external time being a temporal sequence of flow networks), the sink node would have to be re-integrated (have an outflow) to the remainder of the flow network so that the dynamics of the flow network don't dissolve - more generally, it has to become a subgraph that acts as both a source and a sink to the remainder of the network. However, this new ecosystem would also have the universal sink as an absorbing state... It's probably a stretch, but this has a nice interpretation in terms of the diversification of losses into inputs to new nodes - new niches, new connections, new flows. A further stretch, but one I quite like, is that the fact of the non-dissolution of a flow network necessitates the diversification of its losses. More prosaically, it will be forced to recycle in new ways.

    Can draw diagrams if someone is interested.
  • apokrisis
    7.3k
    Which possibilities are closed?fdrake

    The possibility of something else having happened. The existence of the oak is a constraint on the existence of other trees, shrubs, weeds, that might have been the case without its shade. Without the oak, those other entropifiers were possible.

    In what sense was that a mysterious statement? I'd just said "Canopy succession is an example. Once a mighty oak has grown to fill a gap, it shades out the competition."

    So excuse me for being baffled at your professed bafflements in this discussion. I mean, really?

    What degrees of freedom does this create?fdrake

    Again, you claim that I'm hand-waving and opaque, but just read the damn words and understand them in a normal fashion.

    "The mighty oak then itself becomes a stable context for a host of smaller stable niches. The crumbs off its feeding table are a rain of degrees of freedom that can be spent by the fleas that live on the fleas."

    So the oak becomes the dominant organism. And as such, it itself can be host to an ecology of species dependent on its existence. Like squirrels and jackdaws that depend on its falling acorns. Or the various specialists pests, and their own specialist parasites, that depend on the sap or tissue. Like all the leaf litter organisms that are adapted to whatever is particular to an annual rain of oak leaves.

    This is literally ecology 101. The oak trophic network is the primary school level example. You can pick away at its legitimacy with your pedantry all you like, but pay attention to the context here. This is a forum where even primary school science is a stretch for many. I'm involved in enough academic-strength discussion boards to satisfy any urge for a highly technical discussion. But the prime reason for sticking around here is to practice breaking down some really difficult ideas to the level of easy popularisation.

    It's fun, it's professionally useful, I enjoy it. I agree that mostly it fails. But again that seems more a function of context. PF is just that kind of place where there is an irrational hostility to any actual attempt to "tell it right".

    So bear in mind that I use the most simplified descriptions to get across some of the most subtle known ideas. This is not an accident or a sign of stupidity. And an expectation of failure is built in. This is just an anonymous sandbox of no account. My posts don't actually have to pass peer review. I don't have to worry about getting every tiny fact right because there are thousands ready to pounce on faint errors of emphasis as I do in my everyday working life.

    So it is fine that you want that more technical discussion. But the details of your concerns don't particularly light my fire. If you are talking about ecologies as dissipative structures, then I'm interested. If you are talking about something else, like measuring species diversity, or the difficulties of actually measuring exergy/entropy flows in ecosystems, then I really couldn't care less.

    For me. diversity just falls out of a higher level understanding of statistical attractors - https://arxiv.org/abs/0906.3507

    While actually measuring network flows is a vain dream from a metaphysical viewpoint. Of course, we might well achieve pragmatic approximations - enough for some ecological scientist to file an environmental report that ticks the legal requirement on some planning consent. But my interest is in the metaphysical arguments over why ecology is one of the "dismal sciences" - not as dismal as economics or political science, but plagued by the same inflated claims of mathematical exactness.

    What degrees of freedom does this create? How does it create rather than destroy them? How do those degrees of freedom get turned into degrees of freedom for certain organisms? Which organisms? What properties do those recipient organisms have? How do the 'degrees of freedom' in the 'crumbs' relate to the 'smaller stable niches', in what manner do they 'rain'? In what manner are they 'spent'? How does one set of degrees of freedom in the canopy become externalised as a potential for the ecosystem by its deconstruction and then re-internalised in terms of a flow diversification?fdrake

    OK. Degrees of freedom is a tricky concept as it just is abstract and ambiguous. However I did try to define it metaphysically for you. As usual, you just ignore my explanations and plough on.

    But anyway, the standard mechanical definition is that it is the number of independent parameters that define a (mechanical) configuration. So it is a count of the number of possibilities for an action in a direction. A zero-d particle in 3-space obviously has its three orthogonal or independent translational degrees of freedom, and three rotational ones. There are six directions of symmetry that could be considered energetically broken. The state of the particle can be completely specified by a constraining measurement that places it to a position in this coordinate system.

    So how do degrees of freedom relate to Shannon or Gibbs entropy, let alone exergy or non-equilibrium structure? The mechanical view just treats them as absolute boundary conditions. They are the fixed furniture of any further play of energetics or probabilities. The parameters may as well be the work of the hand of God from the mechanical point of view.

    My approach, following from Peirce, systems science, holism, organicism and other -isms stressing the four causes/immanently self-organising view, seeks to make better metaphysical sense of the situation.

    So I say degrees of freedom are emergent from the development of global constraints. And to allow that, you need the further ontic category or distinction of the vague~crisp. In the beginning, there is Peircean vagueness, firstness or indeterminism. Then ontic structure emerges as a way to dissipate ... vagueness. (So beyond the mechanical notion of entropy dissipation, I am edging towards an organic model of vagueness dissipation - ie: pansemiosis, a way off the chart speculative venture of course. :) )

    Anyway, when I talk about degrees of freedom, my own interests are always at the back of my mind. I am having to balance the everyday mechanical usage with the more liberal organic sense that I also want to convey. I agree this is likely confusing. But hey, its only the PF sandbox. No-one else takes actual metaphysics seriously.

    So organically, a degree of freedom is an action with a direction that has to emerge for some holistic or contextual good reason in the physical universe. So why these basic translational and rotational freedoms? Well Noether's theorem and relativity principles account for why actions in these two directions can never be constrained away even in a spatiotemporal system that represents a state of maximal constraint. They can't be parameterised out of existence. Quantum uncertainty has sure rammed that message home now.

    So an ontology of constraints - like for instance the many "flow network" approaches of loop quantum gravity - says that constraints encounter their own limits. Freedoms (like the Newtonian inertias) are irreducible because contraints can make reality only so simple - or only so mechanically and atomistically determined. This is in fact a theorem of network theory. All more complicated networks can be reduced to a 3-connection, but no simpler.

    So in the background of my organic metaphysics is this critical fact. Reality hovers just above nothingness with an irreducible 3D structure that represents the point where constraints can achieve no further constraint and so absolute freedoms then emerge. This is nature's most general principle. Yes, we might then cash it out with all kinds of more specific "entropy" models. But forgive me if I have little interest in the many piffling applications. My eyes are focused on the deep metaphysical generality. Why settle for anything less?

    Now back to your tedious demand that I explain ecology 101 trophic networks with sufficient technical precision to be the exact kind of description that you would choose to use - one that would pass peer review in your line of work.

    Well, again I'm thinking fuck that. I really don't care beyond the possibility that the discussion might be another little window into my bigger picture. I last talked with Ulanowicz probably 15 years ago. So it was interesting to read his recent papers and see how much he has continued on post-retirement in a rather Peircean vein like the rest of that particular crew. (Pattee was the funny one. He got grumpy and went silent for a number of years, despite being the sharpest blade. Then came back blazing as a born-again biosemiotician. The boss once more.)

    Anyhow, fill in the blanks yourself. When I talk of a rain of degrees of freedom, as I clarified previously, I'm talking of the exergy that other entropy degraders can learn how to mine in all the material that the oak so heedlessly discards or can afford to be diverted.

    The oak needs to produce sap for its own reasons. That highly exergetic matter - a concentrated goodness - then can act as a steep entropy gradient for any critters nimble enough to colonise it. Likewise, the oak produces many more acorns than required to replicate, it drops its leaves every years, it sheds the occasional limb due to inevitable accidents. It rains various forms of concentrated goodness on the fauna and flora below.

    Is exergy a degree of freedom? Is entropy a degree of freedom? Is information a degree of freedom?

    Surely by now you can work out that a degree of freedom is just the claim to be able to measure an action with a direction that is of some theoretical interest. The generality is the metaphysical claim to be able to count "something" that is a definite and atomistic action with a direction in terms of some measurement context. We then have a variety of such contexts that seem to have enough of your "validity" to be grouped under notions like "work", or "disorder", or "uncertainty".

    So "degree of freedom" is a placeholder for all atomistic measurements. I employ it to point to the very fact that this epistemic claim is being made - that the world can be measured with sufficient exactness (an exactness that can only be the case if bolstered by an equally presumptuous use of the principle of indifference).

    Then degree of freedom, in the context of ecological accounts of nature, does get particularised in its various ways. Some somewhat deluded folk might treat species counts or other superficialities as "fundamental" things to measure. But even when founding ecology more securely in a thermodynamical science, the acts of measurement that "degrees of freedoms" represent could be metaphysically understood as talking about notions of work, of disorder, of uncertainty. Ordinary language descriptions that suddenly make these different metrics seem much less formally related perhaps.

    That is the reason I also seek to bring in semiosis to fix the situation. You complain I always assimilate every discussion to semiotics. But that is just because it is the metaphysical answer to everything. It is the totalising discourse. Get used to it.

    So here, the key is the epistemic cut that can connect entropy and information. Disorder and uncertainty can be physically related in terms of rate-dependent dynamics and rate-independent information. I earlier linked to a long post which explained how this is currently being cashed out in a big way in biophysics - hence my mention of ATP as the unit of currency that puts a material scale on a cell's metabolic degrees of freedom. ATP is the concentrated goodness of "pure work". We can ground exergy at life's nanoscale, quasi-classical, intersection where a set of entropies just happen with remarkable convenience to converge.

    (Like SX, you really need to add Hoffman's Life's Ratchet to your reading list.)

    Right. I'm sure you will have a bunch of nit-picking pedantry welling up inside of you so I will leave off there. Just remember that I really am engaged in a broad metaphysical project. The correct definition of degrees of freedom is certainly a central concern as it is at the heart of the scientific method. It encodes whatever it is that we might mean by our ability to measure the "real facts" of the world. So it is at this level we can hope to discover the presumptions built into any resulting umwelt or worldview.

    You keep demanding that I cash out concepts in your deeply entrenched notions of reality. I keep replying that it is entrenched notions of reality that I seek to expose. We really are at odds. But then look around. This is a philosophy forum. Or a "philosophy" forum at least. Or a philosophy sandbox even. What it ain't is a peer review biometrics journal.
  • fdrake
    6.6k
    @apokrisis

    Will reply more later, it's late and I'm hooked into some wires.

    Now back to your tedious demand that I explain ecology 101 trophic networks with sufficient technical precision to be the exact kind of description that you would choose to use - one that would pass peer review in your line of work.

    I didn't want you to engage in some kind of organic/mechanical translation exercise, I wanted you to give some specifics of how your concepts act in an ecosystem (or a classical representation of one). Which you did, along with giving an interesting reference. I wanted you to be technically precise with your use of terms - good that you did this.

    I'm sure you will have a bunch of nit-picking pedantry welling up inside of you so I will leave off there.

    I certainly have more sympathy for your view when you do attempt to cash it out in the example. I'm not just looking for why you're wrong, I'm looking for how you could be right.
  • apokrisis
    7.3k
    I wanted you to be technically precise with your use of terms - good that you did this.fdrake

    Great that you think that. But again, my goal is to be technically precise at the most general metaphysical or qualitative level. If you don't yet accept the validity of that, then I don't care. However, just as you insist I cash out my generality in terms of your particular paradigmatic specificity, I only wish you would make an effort to ground your demand for specificity in some more considered ontic basis in fair exchange.

    Why do I have to come all the way over to you? Why do you get to be the judge of "what's good enough" here? And don't pretend that this isn't the rhetorical trap you have sought to establish in this thread.

    I'm familiar with all those kinds of tricks, so if you truly want a deeper level of mutual engagement, you might want to reconsider. I'm not sure that you actually have that much to offer me in return. But we will see if you can eventually pull some metaphysical insights out of the bag with a high surprisal.
  • apokrisis
    7.3k
    (Empiricism and Subjectivity); I think this is exactly the model that ought to be adopted.StreetlightX

    Why adopt the sterile old approach where one side of a dialectical relation must be “wrong” so the other can be “rightl?

    Functional systems - as in political, economic, social, ecological - are the product of their complementary tensions. Both defining aspects of their dynamics are “right” - at least to the degree that they are together in a functional balance.

    So with a social system, that is why both localised or bottom up competition is to be encouraged as much as global top down cooperation. Both need to be vigourous actions even if also then in some mutually beneficial balance.

    You lapse into an us vs them rhetorical mode without even thinking. So that leaves you unable to put something like individualism - localised competitive striving - in its appropriate context. One minute you are against neoliberal selfishness, the next pro PC pluralism. Hence your political position ends up radically confused.

    So yeah, it would be a problem if the institutional level of social order were somehow taken as the positive, the social or personal as the negative. But then the reversal of this prioritisation is just as bad.

    The trick is to see the positive aspect in both the global constraints, the social instinct towards cooperativity, and the local degrees of freedom, the social instinct that celebrates individuality, spontaneity and general striving.

    Well, I say trick. It could also be said to be the bleeding obvious.
  • Deleteduserrc
    2.8k
    I appreciate the conceptually sterilized hystericism you've imputed to me, but I think I'm hysterical in a far more vulgar way, anxiety-at-the-dinner-party. I'm confused, I'm listening to a bunch of people talking, feeling like I don't know if any of it adds up to anything, so 'acting out' if only to get a response. That's where the hysteric's nervous laughter comes from. i got on a public bus, once, years ago and the busdriver complimented me on my sweater, then another dude too. A woman on the bus (with way nicer clothes than me or anyone else there) lost it, crying laughing. I read it like this: "Wtf is this discourse about nice sweaters?! This fucking kid is wearing H&M for christs sake"

    I'm certainly not suggesting you have some master plan to justify x or y and here i am trying to sniff it out. You know critical theory, justifications are sneakier than that. Vanishing mediators. Bam crash! you realized you always had the concepts to justify where you are now, if you follow.

    I don't think theres anything crazy about implying conservatism one moment and neoliberalism the next. thats basically the gop in a nutshell. But take the model and apply to whatever x.

    Anyway im a hysterical bundle of rage and incredulity on a bus of leftists talking ecosystems - ill take it on the chin - and im laughing/crying confused, asking: do you actually believe what youre saying? Whats your career track look like? Is that a zizek sweater from 2006?

    Its fantastic youre a bricoleur but that doesn't mean anything. Ed Gein was (literally) a bricoleur - its politically neutral - it just means you use what you got. For what? For what????
  • Streetlight
    9.1k
    For what? For what????csalisbury

    Joy, OCD.
  • Deleteduserrc
    2.8k
    @StreetlightX it looks bad arguing against joy, so that lets [everyone ever] off the hook. And if a need for control can be chalked up vaguely to a mental illness (that im diagnosed with!) far be it for me to complain. Hope you mean that tho. Full-blooded ocd is a curse. Its not a rhetorical strategy.
  • apokrisis
    7.3k
    Is that a zizek sweater from 2006?csalisbury

    Ed Gein was (literally) a bricoleurcsalisbury

    Glad to see you around again. And in inspired form. Zing!
  • Deleteduserrc
    2.8k
    @apokrisisi zinged you too in my last post but you may have missed it. Listen: i dont trust anyone - but my girl *romantic kiss*
  • Streetlight
    9.1k
    I dunno, I quite like wielding joy aggressively. It's a much underrated way of approaching things - very few know how to deal with weaponized joy. It also means I really don't have much to offer you, by way of your hysteria - I'm bound to disappoint your drives, and I'm perfectly okay with that. I like your provocations though qua provocations- they sharpen me, and are relievingly unformulaic, unlike certain others around here.
  • fdrake
    6.6k
    The possibility of something else having happened. The existence of the oak is a constraint on the existence of other trees, shrubs, weeds, that might have been the case without its shade. Without the oak, those other entropifiers were possible.

    So excuse me for being baffled at your professed bafflements in this discussion. I mean, really?

    I didn't doubt that you understood the 'ecology 101' folklore of how biomass flows and how niches are distributed in the canopy-forest floor trophic network. Why I asked was to see how you used your dictionary of concepts to explain the trophic network in those terms.

    Again, you claim that I'm hand-waving and opaque, but just read the damn words and understand them in a normal fashion.

    So the oak becomes the dominant organism. And as such, it itself can be host to an ecology of species dependent on its existence. Like squirrels and jackdaws that depend on its falling acorns. Or the various specialists pests, and their own specialist parasites, that depend on the sap or tissue. Like all the leaf litter organisms that are adapted to whatever is particular to an annual rain of oak leaves.

    The oak trophic network is the primary school level example. You can pick away at its legitimacy with your pedantry all you like, but pay attention to the context here. This is a forum where even primary school science is a stretch for many. I'm involved in enough academic-strength discussion boards to satisfy any urge for a highly technical discussion. But the prime reason for sticking around here is to practice breaking down some really difficult ideas to the level of easy popularisation.

    I'm not in the business of asking you to describe a simplified trophic network in the usual way it's described then saying 'aha, it was too simple', that'd be an empty rhetorical strategy. Again, what I wanted you to do was use your concepts in a way which clarified their meaning in a simplified trophic network. I take it you agree that a generalised theory of entropy has to be able to instantiate to real world examples, otherwise it's a metaphysics divorced from the reality it concerns.

    It's fun, it's professionally useful, I enjoy it. I agree that mostly it fails. But again that seems more a function of context. PF is just that kind of place where there is an irrational hostility to any actual attempt to "tell it right".

    I thought my responses were precisely demands to 'tell it right' from your perspective. This is commensurate with when you say:

    So bear in mind that I use the most simplified descriptions to get across some of the most subtle known ideas. This is not an accident or a sign of stupidity. And an expectation of failure is built in. This is just an anonymous sandbox of no account. My posts don't actually have to pass peer review. I don't have to worry about getting every tiny fact right because there are thousands ready to pounce on faint errors of emphasis as I do in my everyday working life.

    I'm not in the business of playing peer-review level criticism to your ideas, I don't think my comments have been like that.

    So it is fine that you want that more technical discussion. But the details of your concerns don't particularly light my fire. If you are talking about ecologies as dissipative structures, then I'm interested.

    More technical discussion = apo specifies what his terms mean and how they work in the contexts he describes. I think you'll agree that the style of the post I'm currently replying to is quite different from your usual subsumption of a problem phenomenon to your dictionary of concepts.

    For me. diversity just falls out of a higher level understanding of statistical attractors - https://arxiv.org/abs/0906.3507

    It's an interesting paper. Though it doesn't provide any explicit links between systems that internalise the constraints they use and biodiversity. It looks at specific entropy measures for various spaces then derives maximal entropy distributions subject to constraints. Take the binomial example, it's a discrete distribution with constrained counts, you get out of the analysis in the paper that when you assume a partitioning structure with 2 bins, look at summations of Bernoulli trials - and constrain the mean to a constant - you get the binomial distribution as the maximum entropy one.

    This is a nice link between entropy and the binomial. However, certain configurations of the binomial are entropy maximising - so there's a qualitative distinction between the entropy maximisation occurring in the space of distributions and the entropy maximisation occurring on the maximum entropy distribution that's picked out. Similarly with the space of distributions: the degrees of freedom in the space of distributions are essentially infinite, the degrees of freedom in terms of applied constraints are 1, and the degrees of freedom within the binomial formula are also 1 since the sum is constrained.

    This goes some way to addressing the 'transduction of entropy'. Through a single calculation you end up with the relation of two different entropy concepts and three different degrees of freedom concepts. The caveat is the application of the Lagrange-constraints narrows the application of the results to pre-specified parameter spaces, so an initial justification that a system cares about those constraints (and cares about entropy maximisation) would have to be provided.

    While actually measuring network flows is a vain dream from a metaphysical viewpoint. Of course, we might well achieve pragmatic approximations - enough for some ecological scientist to file an environmental report that ticks the legal requirement on some planning consent. But my interest is in the metaphysical arguments over why ecology is one of the "dismal sciences" - not as dismal as economics or political science, but plagued by the same inflated claims of mathematical exactness.

    Inflated claims of mathematical exactness are a problem across any science whose subject matter is difficult in an epistemic sense. The empirical humanities, including medicine, are actually waking up to this fact at the minute, see the replication crisis.

    OK. Degrees of freedom is a tricky concept as it just is abstract and ambiguous. However I did try to define it metaphysically for you. As usual, you just ignore my explanations and plough on.

    But anyway, the standard mechanical definition is that it is the number of independent parameters that define a (mechanical) configuration. So it is a count of the number of possibilities for an action in a direction. A zero-d particle in 3-space obviously has its three orthogonal or independent translational degrees of freedom, and three rotational ones. There are six directions of symmetry that could be considered energetically broken. The state of the particle can be completely specified by a constraining measurement that places it to a position in this coordinate system.

    So how do degrees of freedom relate to Shannon or Gibbs entropy, let alone exergy or non-equilibrium structure? The mechanical view just treats them as absolute boundary conditions. They are the fixed furniture of any further play of energetics or probabilities.

    I'm not sure what you mean by boundary conditions, but I'm guessing it's something like 'background assumptions required for the formation of a measure'.

    The parameters may as well be the work of the hand of God from the mechanical point of view.

    I appreciate that you are attempting to find a sense of 'becoming relevant' of parameters, and I think the paper you linked about maximum entropy distributions is a step in the right direction. But I don't think it's appropriate to treat parameters as 'God given', as you put it.

    If you want to mathematise something, it'll have a bunch of assumptions of irrelevance so that it fits on a page. EG, when you look at something solely in terms of a binomial distribution, you care about counts of stuff - not how the counts became relevant. A phenomenological description of what's happening in a system is always useful and should be a mandatory preparatory measure for a couple of reasons. Maybe you'll see some dialectical correspondence in this:

    (1) It expresses the model building intuitions and the purported significance of included terms and the irrelevance of excluded ones.
    (2) It allows the relation of the mathematisation to the imaginative background of the phenomenology that derived it.

    So I say degrees of freedom are emergent from the development of global constraints. And to allow that, you need the further ontic category or distinction of the vague~crisp. In the beginning, there is Peircean vagueness, firstness or indeterminism. Then ontic structure emerges as a way to dissipate ... vagueness. (So beyond the mechanical notion of entropy dissipation, I am edging towards an organic model of vagueness dissipation - ie: pansemiosis, a way off the chart speculative venture of course. :) )

    Anyhow, fill in the blanks yourself. When I talk of a rain of degrees of freedom, as I clarified previously, I'm talking of the exergy that other entropy degraders can learn how to mine in all the material that the oak so heedlessly discards or can afford to be diverted.

    The oak needs to produce sap for its own reasons. That highly exergetic matter - a concentrated goodness - then can act as a steep entropy gradient for any critters nimble enough to colonise it. Likewise, the oak produces many more acorns than required to replicate, it drops its leaves every years, it sheds the occasional limb due to inevitable accidents. It rains various forms of concentrated goodness on the fauna and flora below.

    Instantiating it:

    Oak community has X number of species dependent solely on its existence to exist.
    Oak community has Y number of species which are reduced in number solely from what would happen without the oak community.

    These are degrees of freedom in the first sense.

    Species in X have network of flows. Oaks removed, X goes to 0.
    Species in Y have networks of flows. Oaks removed, Y probabilistically increases.

    Flows:

    Complete degradation of network consisting of X, inputs to X are reassigned to other networks.
    Total throughput in Y increases if Y has species which were constrained by species in X - since input node to Y increases if it is a function of input to X.

    Total throughput - sum like variable - assumed constant so long as trophic network is stable or permits immediate recolonisation of destroyed niches with the same efficiency and that concentration of flows will not degrade the ecosystem - decreasing degrees of freedom in the first sense.

    If energy from removal of X's effects are distributed evenly among functional roles, degrees of freedom in the second sense increase a lot. If they are equally concentrated, degrees of freedom remain roughly constant. Degrees of freedom in the second sense - similar to exponentiation of flow entropy.

    Measurement - variables
    X and Y can be identified without error, but inclusion in study can miss some out.
    Total throughput - two measurements required to detect change, likely noisy, nodes in study can miss some out.

    Expected behaviour-
    Entropy maximisation - requires that distributional changes resulting from X's removal increase entropy in the functional sense. Occurs through function of total throughput and the proportions obtained of it by new niches.
    Generalised entropy maximisation - has occurred if distributions in the pre-removal of X era are shifted closer to derived maximal entropy distribution with entropy maximising parameters.

    Does this sound like a transcription of the canopy-floor ecosystem into your abstract register?

    If so: there's rather a lot of counterfactuals there. Especially to assume without evidence.

    Anyway, when I talk about degrees of freedom, my own interests are always at the back of my mind. I am having to balance the everyday mechanical usage with the more liberal organic sense that I also want to convey. I agree this is likely confusing. But hey, its only the PF sandbox. No-one else takes actual metaphysics seriously.

    So an ontology of constraints - like for instance the many "flow network" approaches of loop quantum gravity - says that constraints encounter their own limits. Freedoms (like the Newtonian inertias) are irreducible because contraints can make reality only so simple - or only so mechanically and atomistically determined. This is in fact a theorem of network theory. All more complicated networks can be reduced to a 3-connection, but no simpler.

    So in the background of my organic metaphysics is this critical fact. Reality hovers just above nothingness with an irreducible 3D structure that represents the point where constraints can achieve no further constraint and so absolute freedoms then emerge. This is nature's most general principle. Yes, we might then cash it out with all kinds of more specific "entropy" models. But forgive me if I have little interest in the many piffling applications. My eyes are focused on the deep metaphysical generality. Why settle for anything less?

    Looking at the how your background conceptions apply to the real world is an excellent way of revealing conceptual and practical problems in your metaphysics. It isn't settling for less

    Surely by now you can work out that a degree of freedom is just the claim to be able to measure an action with a direction that is of some theoretical interest. The generality is the metaphysical claim to be able to count "something" that is a definite and atomistic action with a direction in terms of some measurement context. We then have a variety of such contexts that seem to have enough of your "validity" to be grouped under notions like "work", or "disorder", or "uncertainty".

    So "degree of freedom" is a placeholder for all atomistic measurements. I employ it to point to the very fact that this epistemic claim is being made - that the world can be measured with sufficient exactness (an exactness that can only be the case if bolstered by an equally presumptuous use of the principle of indifference).

    Hurrah, it was a placeholder. I understood what you meant!

    Then degree of freedom, in the context of ecological accounts of nature, does get particularised in its various ways. Some somewhat deluded folk might treat species counts or other superficialities as "fundamental" things to measure. But even when founding ecology more securely in a thermodynamical science, the acts of measurement that "degrees of freedoms" represent could be metaphysically understood as talking about notions of work, of disorder, of uncertainty. Ordinary language descriptions that suddenly make these different metrics seem much less formally related perhaps.

    Could you comment on my attempt at instantiating your concepts to the canopy-floor ecosystem example?

    That is the reason I also seek to bring in semiosis to fix the situation. You complain I always assimilate every discussion to semiotics. But that is just because it is the metaphysical answer to everything. It is the totalising discourse. Get used to it.

    Why do you think semiotics is the totalising discourse? I'm quite suspicious of the claim that there are genuine totalising discourses; attempts to reduce reality to one type of thing fail for precisely the same reasons systems science became so popular (perhaps with some irony resulting from the view of everything as a system).

    You keep demanding that I cash out concepts in your deeply entrenched notions of reality. I keep replying that it is entrenched notions of reality that I seek to expose. We really are at odds. But then look around. This is a philosophy forum. Or a "philosophy" forum at least. Or a philosophy sandbox even. What it ain't is a peer review biometrics journal.

    What kind of description would satisfy your desire for a better 'ontic development' of my presumptions?

    You keep complaining that I'm attacking your concepts because solely they're not biometrically sound. This is the same kind of thing as saying that I have a mechanist's vantage point on ecology. The reason I'm using pre-developed entropy measures is to highlight the ambiguity in your presentations of the concept. The purpose was to get you to describe how stuff worked in your view without the analogising.

    So I can add to my apokrisis dictionary: what's a vague-crisp distinction when it's at home? And what's the epistemic cut?
  • fdrake
    6.6k
    @apokrisis

    I'm not going to respond to anything quantum or differential-geometric unless you think it's essential. Things are involved enough as it is.
  • apokrisis
    7.3k
    Sorry fdrake, but I don't get where this is going. Your responses are vague as if you are only intent on creating some endless descent into technicalities with no finishing line. If you don't signal what you agree with, then I'm just guessing at where any useful disagreement lies.

    Do you want to have a go at summing up what you think has been revealed to be essentially wrong about my general metaphysical approach here? What would be the core disagreement in terms of orientation?

    I explained for instance that a degree of freedom is a placeholder for the brute claim to be able to measure "actions with directions". You replied, hurrah, you were right that it is a placeholder. But then didn't comment at all on the kind of placeholder I said it was.

    Then again, I specified that we find various notions of "actions with directions" being counted. Degrees of freedoms can be decomposed into various more qualitative or contextual notions, like "work", "disorder", "uncertainty". Once more, no comment whether you either agree or disagree.

    Nor will you tie anything back to my original reply to the OP - my mention of Salthe/Ulanowicz's lifecycle analysis and its applicability to political theory. A metric like ascendency tries to pick up on something even more subtle than the usual dissipative structure story.

    Degrees of freedom in this context are the reserve, the overhead, that a living system needs to keep in reserve so as to be able to adapt to perturbation. An organism (or society) can't afford to spend all its entropic "income" on here and now maximal growth. It wants a reserve of fat, a reserve of degrees of freedom, to deal with unexpected challenges.

    My point is that "degrees of freedom" is a useful generic term because it is dichotomous to "constraints", it signals "whatever is definitely countable in terms of some parameterised theory", and it is undefined enough to encompass an ever branching family of thermodynamically related thought - as in capturing this notion of a reserve of adaptive capacity. So I don't use terminology in some unthinking handwaving fashion, as has been your repeated accusation. There is a proper metaphysical structure that organises my ideas. And it is a way of looking at the issues which I learnt firsthand from folk like Salthe and Ulanowicz.

    So again, is there anything more here than you want me to break a still-developing metaphysics of a pansemiotic Cosmos down into everyday measures you can employ to do a better job of modelling some ecosystem with?

    Sum it up. What do you agree with and what is any core disagreement or vital question that remains to be tackled?
  • charleton
    1.2k
    Interesting reaction. I think you could put yourself in danger of seeing the results of the dynamics which are actually contingent upon unique conditions are elevating them into a a series of casual factors. The tail of the system waging the dog of necessity.
bold
italic
underline
strike
code
quote
ulist
image
url
mention
reveal
youtube
tweet
Add a Comment

Welcome to The Philosophy Forum!

Get involved in philosophical discussions about knowledge, truth, language, consciousness, science, politics, religion, logic and mathematics, art, history, and lots more. No ads, no clutter, and very little agreement — just fascinating conversations.