24 May 2012

Ontology for Whose Sake?

Filed under: Psychology — ktismatics @ 11:54 am

On Tuesday I was browsing the 2 March 2012 Times Literary Supplement when I came across an article written by Daniel Dennett. In the essay, entitled “Sakes and Dints and Other Definitions that Philosophers Really Need Not Seek,” Dennett wonders about the ontological status of numbers, melodies, miles, electronic dollars, cahoots, smithereens, voices, haircuts, centers of gravity, minds — “the riotous assembly of candidate things” that humans deal with all the time. Should all of them be eliminated from material reality as mere epiphenomena, or should there be some robust ontology of objects that encompasses them all?

Ontology, it seems to me, has been resolutely reductive and minimalist in its attempts to come up with an exhaustive list of kingdoms, classes, genera and species of things. No doubt the motivation has been the right one: to prepare the dishevelled cornucopia for scientific accounting, with everything put ultimately in terms of atoms and the void, space-time points, or (by somewhat different reductive trajectories) substances and universals, events and properties and relations. As is well known, these Procrustean beds provide metaphysicians with no end of difficult challenges, trying to fit all the candidates into one austere collection of pigeonholes or another.

Take holes for example:

Our reliance on the concept of holes has a deep biological or ecological source, since holes are Gibsonian affordances par excellence, put to all manner of uses in the course of staying alive and well in a hostile world. Ubiquitous and familiar though they are, it is surprisingly hard to say what holes are made of, if anything, what their identity conditions are, whether they are concrete or abstract, how to count them, and so forth.

Dennett tentatively suggests regarding the hole as a meaningful category in naive physics but not in scientific physics.

But not so fast: holes may play a potent role in organizing the patterns studied in the special sciences — consider the importance of membranes and their multiple varieties of portals in cell biology, for instance — and what about the slit that figures so prominently in quantum physics?

Dennett regards the categorization of things as “more a matter of diplomacy than of philosophy.”

It is not that there is or must be — there might be — a univocal, all-in, metaphysical truth about what there is, but just better and worse ways of helping people move between different ontological frameworks, appreciating at least the main complexities of the failures of registration that are encountered.

He concludes:

The perspective I would recommend is that of the diplomatic anthropologist, not the metaphysician intent on limning the ultimate structure of reality. The ontology of everyday life is now teeming with items that sit rather awkwardly in the world of atoms and molecules. If we can understand how this population explosion came about, and why it is so valuable to us as agents in the world, we can perhaps discharge our philosophical obligations without ever answering the ultimate ontological question. To me it looks more and more like professional make-work, an artefact of our reasonable but ultimately optional desire for systematicity, rather than a deeper mystery in need of solving.



  1. D.D. is being whimsical. The classical subject matter of ontology is being as such. Such questions as ‘what is it for something to be something’. Is a substance something of which x, y, z is said but it is not said of anything i.e.the ultimate subject of predication. All these questions which are of no interest to D.D. which are probably not important because he is not interested in them. He’s a bright, so that’s all right.

    Comment by ombhurbhuva — 24 May 2012 @ 1:56 pm

  2. Dennett is often whimsical. I think he’s also being pragmatic: what are all of these things for? Ontologies are also a kind of thing, of which there are many that have varying uses. More than once I’ve seen Dennett shoveled into the same hole as the eliminative materialists; here he explicitly resists that definitive categorization. For some purposes I’m that, he acknowledges; for other purposes, not. He’s also been interpreted, wrongly in my view, as denying intentionality, which seems a prerequisite for pragmatism. He certainly accepts the idea of a reason that something happens; e.g., via natural selection. What about reasons for doing something? He acknowledges intentionality explicitly at the end of the article: “…and why it is so valuable to us as agents in the world…”

    Comment by ktismatics — 24 May 2012 @ 2:28 pm

  3. Here’s a post from a year ago and a linked video of Dennett on reasons why and reasons for.

    Comment by ktismatics — 24 May 2012 @ 7:34 pm

  4. I appreciate what Dennett is saying, but I tentatively disagree with him. A little. Maybe.

    A melody does sit well, at least in one sense, with scientific physics. It’s the conceptual sense of a melody that doesn’t.

    I think the way we do ontology has not been very effective. Giving ontological status to “electronic cash” has an almost built-in Platonism, which leads to an almost inescapable sort of misapprehension. But I’d really love to see a pragmatic, physicalist ontology that includes mental phenomena — or at least know what it might look like.

    I think the “hard problem of consciousness” is hard because the closest thing that it’s like (a computational machine) is still very different from it. We lack billiard balls for the mind. To make things worse, thoughts and ideas have a sort of physical existence that is radically different from our central concept of physical things. So it’s hard to think about them, systematize them or model them.

    There are so many built-in problems that cahoots will probably always sit poorly with molecules. But… if we’re more nominalist about it, and we try to figure out how the physical existence of a thought works – how and why it persists; how it is retrieved, manipulated and transformed; how it can represent or refer to something – we can really begin to see much of philosophical ontology as just “a way of speaking about things”.

    Comment by Asher Kay — 25 May 2012 @ 9:40 pm

  5. I remember having an extended discussion with Kvond about the reality of songs — this was back when a number of us were trying to engage with Levi’s ontology directly, with mutual commenting on one another’s blogs and so on. Looking back on those posts I was contending that a song is (usingThomasson’s term) an “abstract object,” characterized as a structured pattern of information that isn’t restricted to any particular space-time coordinates and that can manifest itself materially in a variety of ways: in a voice or a musical instrument, on a piece of paper, or in someone’s brain. But can it be a song if it doesn’t become concrete by being played, sung, whistled, etc.? It’s the same issue with ideas, words, jokes — things that Dawkins clumped into the category of memes. I recall that Kvond didn’t buy the idea of an abstract object existing as an abstract pattern of information. For him there was no “Dixie”: there was only this or that particular performance of Dixie. This I guess was Spinozist thinking, and would probably also be Latourian: this performance of Dixie on a harmonica is ontologically different from whistling Dixie is different from a tape recording of the harmonica performance and so on. You brought up the reality of furriness recently: another abstract category that could conceivably be described in physical terms but that remains nonetheless an abstract category. I wonder whatever happened to Kvond…

    Earlier today I read this in the chapter on Self in Deacon’s Incomplete Nature:

    “Mental categories aren’t merely passive reflections of the world; they exist to structure adaptation to the world. For this reason, the mere resemblance of an object to a perceptual class can be what causes that object to be modified in a particular way by an animal or person… Minds capable of symbolic references can literally bring even the most Platonic of conceptions of abstract forms into the realm of causal particulars. To list some extreme but familiar examples, a highly abstract concept like artistic beauty can be the cause of the production of vastly many chiseled marble analogues of the human female form; a concept like justice can determine the restriction of movement of diverse individuals deemed criminal because of only vaguely related behaviors each has produced; and a concept like money can mediate the organization of the vastly complex flows of materials and energy, objects, and people, from place to place within a continent. These abstract generals unquestionably have both specific and general physical consequences. So human minds can literally transform arbitrarily created abstract general features into causally efficacious specific physical events.”

    This too is a pragmatic sort of ontology. That makes it correlationist of course, but Deacon is explicit in arguing that information isn’t a thing or an energy but a relational aboutness characterized by constraints, which is where what you call the conceptual aspect of melodic information comes in. I found it interesting that Dennett spent so much time on the ontological status of holes. Deacon emphasizes the importance of lack as a kind of anti-thing; I wonder how much Dennett is influenced by Deacon here. I googled the two names and they have co-authored and commented on various specific ideas related to mind and evolution, so I wouldn’t be surprised if there’s synergy between the two of them. I wonder if Deacon bitched about Dennett not citing him in the holes discussion…

    Comment by ktismatics — 25 May 2012 @ 10:27 pm

  6. I kind of miss Kvond. He knew how to get on my last nerve, but I still think about some of the points he made, which at least indicates that a meaningful discussion occurred.

    Spinoza was a nominalist in the sense I mean (although I believe people still dispute that). I don’t think I totally agree with Kvond about Dixie. I just think it’s a sort of category error to think about the existence of Dixie the abstract song as being the same as the existence of particular Dixie-like patterns. Unless we’re talking about Dixie the abstract song as a particular instance of someone thinking about the abstract concept.

    Deacon sprung to mind for me too with the holes thing. I wonder if Dennett was indirectly commenting on Deacon.

    I’m still working at formulating the ontological distinction I want to make. It’s still a little fuzzy to me. The basic idea is that it would start with the notion that we create mental models to explain the world, and that those models operate in a particular way and do particular things. If we come up with an explanatory model that explains someone composing a song in her head, deciding to do something, imagining a conversation, planning out an action, solving a problem, generalizing particulars, or composing a conceptual model, then we’ve done the explaining that we need to do. Once that happens, we can then be more at peace with the way we speak about melodies and holes. It’s a little like knowing how a magic trick works.

    In a way, that seems like the same thing Dennett is saying.

    Comment by Asher Kay — 26 May 2012 @ 4:32 pm

  7. I meant to ask — did that paragraph from Deacon make sense to you? I kept wanting to re-state it in terms of a neural network processing information.

    I’d call it correlationist, too, but it’s a diffuse and fuzzy sort of correlation.

    Comment by Asher Kay — 26 May 2012 @ 4:43 pm

  8. Harman’s big insight in Tool-Being (which I haven’t read) came from Heidegger’s distinction between “ready to hand” and “present at hand”. The hammer in ready-to-hand mode is a tool for hammering; in present-at-hand mode the same hammer is the same object uninvolved in its pragmatic function as a tool. Harman decided that the same thing can’t be two different objects, so he distinguished between an object’s properties and its essence. The object’s properties participate in its interactions with other objects — e.g., the ready-to-hand hammer’s tool properties interact with the guy who wants to hang a picture in the hallway — while the present-at-hand essence of the hammer withdraws from its use as a hammer.

    I’m far more aligned with your view, Asher. For the guy wanting to hang the picture, things like hammers, nails, rulers, pencils, pictures, and so on participate in the guy’s mental model of picture-hanging. If there is no hammer in the toolbox, a palm-sized rock out in the yard might come to participate in the picture-hanger’s mental model as a ready-to-hand hammering tool. The tool-ness of the hammer is designed and built into it, its ready-to-hand properties being paramount. But the hammer’s pragmatic utility is predicated on some specific properties of the object qua object: its weight, its long narrow handle, the hardness of its striking surface. Do those properties interact with the hammer’s tool-properties? Sure: the tool-use is predicated on the physical properties. So where is the essence? Is it in the molecular composition? Those are properties of use to the scientist, Harman would probably say. So where is the essence of the hammer-object? It has to be non-material.

    After all of the humans have gone extinct from this planet, an alien civilization might show up and start looking around. If they came upon a hammer and didn’t take into account its intended use, the aliens wouldn’t really understand the essence of the hammer within the extinct humans’ “mental model” of the world they had occupied. The aliens might never suspect that that rock had once been used to hang a picture by some guy who couldn’t find his hammer.

    Comment by ktismatics — 27 May 2012 @ 7:20 am

  9. With respect to tunes, I’m fine with accepting this particular performance of Dixie on the harmonica as being a separate thing from someone’s whistled rendition of Dixie. Clearly though they’re both performances of the same song. A performance is a composite, consisting of the song, the performer, the instrument, along with idiosyncratic features of the specific interaction of those three components in playing the song in this particular time and place. The performer and the instrument are material objects; the song is the information pattern of note lengths, pitches, rhythms, etc. This description seems to work fine as long as you don’t go all-in for either materialism, in which only the specific performances are real, or idealism, in which only the abstract song decoupled from specific performances is real.

    Comment by ktismatics — 27 May 2012 @ 9:18 am

  10. Sounds like we are on the same track. If I’m framing a wall, a tack hammer may as well be a screwdriver — it doesn’t have anything like the properties needed to punch a 16d common nail through a 2-by-4.

    To me, “tune” and “hammer” are basically the same kind of concept. Hammers are more oriented toward materiality than tunes, but both a purely physical (and our general concepts of tunes and hammers are also physical, though less material).

    The alien thought-experiment is useful in a couple of ways. The first way is imagining the aliens finding a hammer and trying to “deduce” what it is used for. It’s pretty easy to see that hammers are just as dependent on hands as they are on nails. If the aliens don’t know what we looked like and don’t have hands themselves, they’re going to have a hell of a time figuring a hammer out.

    The second way is realizing that the concept of hammers needs to be physically embodied. If all the people who know what a hammer is die, and the rest of the people try to figure it out, they have an advantage over the aliens, and the concept has a better chance of being “re-discovered”. The concept of a hammer is purely physical. If all the hammers disappear and all the people who have a concept of hammers die, hammers are just plain gone.

    Hammers are easier for us cognitively than Dixie, but I don’t think it’s anything different. If all the people who know Dixie die and all the recordings are gone, Dixie is gone. If someone records a process for producing Dixie (sheet music + meta-description of how sheet music works), Dixie still exists, but latently (this is where constraints are helpful, because the whole exists/doesn’t exist thing feels inadequate, whereas the recorded process as a constraint making the production of Dixie a possibility is more like what is happening).

    What we have to work to get our heads around is that there is no identity in any of these things. Each instance partially overlaps with the others, and each concept is only a partial match to the patterns of the others (further, different people would have different opinions about what constraints are sufficient to produce Dixie). Just as we embody the concept of Dixie in the physical world by recording it (or recording the process for producing it), we embody it in the physical world (specifically, our brains) by having the concept. Ontologically, there’s no difference. It’s easy for philosophers to forget that concepts need to be maintained over time. They forget it because they view concepts as things.

    So from my perspective, you can’t get to any of these things with a purely material view, or a material-plus-organization view. It has to be a process, over time, which basically dismantles the idea of identity. This is where cognitive and information-theoretical views of things shine. Not only can we sometimes figure out Dixie from a noisy recording (error correction), but we can fuzzily classify tunes, and say that a jazz saxophonist “threw a piece of Dixie into that solo” or that a composer varied the rhythm and notes enough that it’s no longer Dixie.

    Comment by Asher Kay — 29 May 2012 @ 9:53 am

  11. As usual, I forgot to throw in something. My wife teases me that I am the “master of using things for something other than their intended purpose”. When I don’t have a particular tool or part, I look around and try to find something that will “do the trick”, usually with some modification. I think this promiscuity of use makes it easier for me cognitively to dispense with identity as being important.

    Comment by Asher Kay — 29 May 2012 @ 9:57 am

  12. “Sounds like we are on the same track.”

    I agree — which means we can at least potentially agree on the relationship between the “track” as an abstract pattern of ideas vis-a-vis our separate ways of walking along that track. I also agree that a tune is like a hammer: both are abstract patterns that can be made material in a variety of ways and that depend on their being “about” something for entities other than themselves; e.g., humans who hammer or whistle.

    I’m not metaphysically savvy enough to understand the connotations of “identity.” I’m prepared to acknowledge that the boulder in the stream bed is different from the water rushing by it and from the gravel on which it rests. It’s only at the interfaces between boulder, water, and gravel that the boundaries get fuzzy, where friction from both water and gravel work at the microlevel to wear away the boulder’s mass bit by bit. The boulder does not actively protect its integrity from erosion though, so maybe in that sense it has no identity.

    Deacon deems it an important feature of self-generating, self-repairing, self-replicating systems or “autogens” that they form and maintain some sort of container that actively buffers the system from its environment. Page 309:

    “This self-reconstituting dynamics provides an active self-similarity-maintaining quality which constitutes a form of individuality, or ‘self,’ that does not otherwise exist outside of living processes. There is both an individual autogenic identity, as a closed, inert, but potentially self-reconstructing unit, and a self-maintaining lineage identity, due to the transmission of relatively invariant intrinsic dynamical constraints and molecular types from generation to generation as a result of replication. Self-reconstitution does not completely maintain material identity across time because it allows for molecular replacement, and it does not maintain energetic or dynamical continuity across time either, since it may persist in a static phase for extended periods. But this self-reconstitution capacity does maintain a persistent and distinctive locus of dynamical organization that maintains self-similarity across time and changing conditions. And yet ultimately there is no material continuity, as autogens are disrupted only to be reconstituted and replicated with newly synthesized components. Only the continuity of the constraints that determine the autogenic causal architecture is maintained across repeated iterations of dissolution and reconstruction.

    “So an autogen has identity only with respect to this persistent general pattern of constraint maintenance and replication, and irrespective of any particular molecular constituents. Indeed, it is the continuity of the inheritance of constraints on its molecular dynamics that constitutes this individuality. But identity may vary as an autogen lineage evolves variant forms of this defining dynamics. For all these reasons, an autogen self and autogen lineage identity are examples of efficacious general types, in a philosophical sense (see chapter 6). To be more specific, an autogen is an empirical type determined only by the continuity of these dynamical constraints, which are themselves expressions of dynamical limitations — potential modes of change not expressed.”

    I’d have to revisit chapter 6 to see what Deacon means by “efficacious general types,” but in the excerpt from chapter 15 I quoted in comment 5 above I presume that he would regard a tune or a hammer as a general type, defined not by its materiality but by its constraints and its efficacy in “being for” something. So a hammer constrains the kinetic energy of a blow to the striking surface; Dixie constrains sonic energy to a particular sequence of notes. Surely Dixie as a general type isn’t an ideal form, as you point out. There are variations, partial matches, statistical distributions in the constraints constituting the fuzzy boundary between Dixie and not-Dixie, between hammer and not-hammer. Deacon might even regard Dixie as having more of an identity than does the boulder in the stream bed.

    ““master of using things for something other than their intended purpose”

    Maybe you should put this as your job title on resumes, business cards, etc., Asher. “Will do the trick”? Maybe not.

    Comment by ktismatics — 29 May 2012 @ 12:14 pm

  13. Yeah, I think any foray into prostitution on my part would be fiscally unsound….

    I’m looking at identity more as “exactly-the-sameness” rather than individuation or separation, but it looks like they’re two sides of the same coin. A process-over-time conception of the autogen or the eddy sees the identity as a maintenance of self-similarity, but one that can’t be static (no precise sameness over time), even for the boulder, which erodes and accretes over time.

    The idea of identity, I think, comes from a weird trick of cognition where we can separate the type from the thing. It leads to a sort of Platonic idea that if we’re all thinking about “the same Dixie”, then there must be some independently existing “form” of Dixie, when actually Dixie is a fuzzy conglomeration of things with separate physical existences, and “identifying” it, is a set-inclusion evaluation. Any time we want to deal with the “form”, we have to subtract it out from something.

    Identity, in short is really classification. And classification is process-like. And viewing it that way allows us to dance between platonic and nominalist views.

    All of this makes me wonder…. If you came across an object that was naturally and randomly ocurring, but which had a lot of hammerlike properties (shape, hardness, etc), would you hesitate to call it a hammer? Or if some weird geothermal feature produced tonal earth exhalations that matched the notes and rhythm of Dixie, would you hesitate in saying that it is Dixie that is being played? You definitely wouldn’t say that the geothermal feature was “for” playing Dixie.

    I am seriously hoping that I don’t get Dixie stuck in my head. I really, really hate that tune.

    Comment by Asher Kay — 29 May 2012 @ 1:24 pm

  14. That’s helpful about “identity.” I tend to think of it as a psychological term, as with the William James excerpt from a few posts ago. James extends identity to other organisms besides humans: “Even the trodden worm, as Lotze somewhere says, contrasts his own suffering self with the whole remaining universe, though he have no clear conception either of himself or of what the universe may be. He is for me a mere part of the world; for him it is I who am the mere part. Each of us dichotomizes the Kosmos in a different place.” James had issues too in defining identity, for organisms and for inanimate objects in the world. Deacon and James are both one degree of separation from CS Peirce in arriving at a pragmatic definition of identity based on “habits” or constraints imposed on the “blooming buzzing confusion.”

    There were plenty of rocks littering the earth long before there were tool-using beings who might have appreciated the hammer-esque affordances of those rocks. So I guess I’d say that hammerness isn’t intrinsic to the object, but is defined relationally and pragmatically with respect to the intention of hammering. This morning I actually heard what I thought was music but that turned out to be water trickling into a metal sink. (Thankfully it did not resolve itself into any recognizable variant of Dixie.) Once I realized the source of the sound I decided that I was mistaken, that it wasn’t music after all. So I guess I intuitively distinguish between music and non-music based on the intentionality of the source. I did find myself trying to hum the tune the water had made though, so I was mentally reframing the unintentionally-generated sound pattern as being “for” music. What is the sound that dripping water makes when there’s no one there to hear it? I think it emits pressure waves of particular amplitudes and frequencies, but unless there’s someone listening those wave patterns can’t become music. In this regard evidently I agree with Deacon: information, including musical information, is relational.

    Comment by ktismatics — 29 May 2012 @ 2:14 pm

  15. I was going over the chapter on constraints for a DV piece I’m working on, and I think I see where Deacon messes me up. It’s a part where he’s examining how to escape the conclusion that general types are epiphenomenal (things merely imposed upon reality by our minds). He says (really hoping blockquote tags work in comments):

    Does this necessarily imply that these similarities and regularities are somehow mere figments of mind? Can they be mental abstractions and physically relevant causal features of the world at the same time? One reason to think that mental abstraction does not imply physical epiphenomenalism is that mental processes are themselves also dependent on physical processes. To even perceive regularity or pattern this act of observation must itself be grounded in a habit of mind, so to speak. In other words, to attribute physical regularity to some perceived or measured phenomenon presumes a prior mental regularity or habit with respect to which the physical regularity is assessed.

    This seems to agree (in a slightly strange way) with my conception of physical, embodied cognition. The insight that’s always turned me on is that the perception of regularity is physically caused by regularity in the external world — in essence, external regularity generates a regularity in the (physical) sensory/perceptive process in the brain. But I was a little taken aback by the “dependent on”. Why “dependent on”? Mental process *are* themselves physical processes. Maybe he didn’t mean anything special by this. But then he goes on to say:

    Even if we grant that general tendencies of mind must already exist in order to posit the existence of general tendencies outside the mind, we still haven’t made any progress toward escaping this conceptual cul-de-sac. This is because comparison and abstraction are not physical processes.

    This seems just plain wrong to me. Comparison and abstraction are very much physical processes. We can construct fairly simple artificial neural networks that perform these tasks.

    Not sure what to think. I’m hoping maybe you have some insight.

    Comment by Asher Kay — 30 May 2012 @ 10:55 am

  16. “mental processes are themselves also dependent on physical processes” — I figured this was Deacon’s refusal of the eliminative materialist stance that reduces mind to brain and thought to neural firings. He’s consistent on that score, predicated largely on his idea of a constraint being not a material something but rather an absence of something. Presumably mental processes result from prior constraints while propagating new ones, all within the substrates of matter and energy on which they depend.

    “general tendencies of mind must already exist in order to posit the existence of general tendencies outside the mind”

    I’ve not read Kant, but that sounds like what I understand as a Kantian notion of a priori categories in the mind. Maybe he’s not being that specific. Certainly it takes a generally clever sort of mind to compare one thing to another, regardless of what those things happen to be. It requires at minimum the ability to call up from memory the representation of some other object or class of objects previously experienced, against which the object currently being perceived can be compared. But then…

    “comparison and abstraction are not physical processes.”

    Why doesn’t he say that, as mental processes, comparison and abstraction depend on physical processes? Again, I have some trouble with Deacon’s idea of constraint being a causal force, a void or incompleteness in the physical world. If it isn’t physical, then what is it?

    I’ll definitely participate in discussing your Vole posts, Asher. I just wanted to write some thoughts while the book was still fresh in my mind and before I have to return it again to the library.

    Comment by ktismatics — 30 May 2012 @ 12:48 pm

  17. I agree about the eliminativist materialist thing. It’s just that he seems to be advancing a distinction between “materialist” (aligned with a things + relations view) and physicalist (aligned with a everything-is-a-process view). So something can have causal potency in a way that doesn’t spring from matter being in a particular configuration, and still be completely physical.

    It does seem sort of Kantian. I guess what gets me is that saying something is not physical is a really big deal when philosophers like McGinn look at it, because it sets up the whole dualism thing where you have to explain how non-physical things can inflect upon physical ones. That’s a long way to stretch the idea of absence, if that’s what is supposed to get around the problem.

    And beyond that, I think there’s a way to describe the whole thing in physical terms which serves the same purpose and still has a big role for absence. I was planning to post about that on Dead Voles, but I don’t want to be mistaken about Deacon’s view on something like this, especially after complaining that people like McGinn haven’t read closely enough.

    I went over the whole thing with my wife, and her suggestion was to just e-mail Deacon about it. That initially seemed unthinkable to me, but I ended up doing it anyway. We’ll see if he responds.

    Comment by Asher Kay — 30 May 2012 @ 9:06 pm

  18. Looks like I gave away my identity there somehow. Lissack should be pleased.

    Comment by Asher Kay — 30 May 2012 @ 9:14 pm

  19. The thing about constraints is that in Deacon’s examples they have physical causes. He’s explicit that the constraint triggering the anti-entropic self-organization comes from outside the system; e.g., the boulder in the stream, the chamber through which the internal combustion explosion is channeled. So it’s not like there’s something mysteriously missing; the source of the constraint can always be directly traced, the absence traced to a presence. And as a consequence I’m not sure how constraint and absence provide the missing factors in understanding the emergence of life.

    Good idea to email Deacon; he seems accessible. Lissack has made it impossible for him to comment on the Vole thread now, so personal contact might do the trick.

    Comment by ktismatics — 30 May 2012 @ 9:53 pm

  20. There’s one place where I know absence is *really* important, and that’s in Deacon’s response to Jaegwon Kim. Basically, Deacon says that yeah, everything is reducible just like you say, but any reduction doesn’t address what’s *not* there. It seems almost like a trick of logic, and it only works if absence can be causal.

    But I’m struck by the same thing you are: as long as absence can be equivalently expressed as not-presence, it’s hard to see how something more is being said.

    One thing that might be key is the idea that when classifying by presence (the presence of a property or attribute), there is a qualitative judgement involved, whereas when classifying by absence, the judgement can be purely quantitative. At least that’s what Deacon says. I’m not totally convinced.

    Comment by Asher Kay — 30 May 2012 @ 10:26 pm

  21. Here’s the quote for the qualitative/quantitative point:

    The general logic is as follows: If not all possible states are realized, variety in the ways things can differ is reduced. Difference is the opposite of similarity. So, for a finite constellation of events or objects, any reduction of difference is an increase in similarity. Similarity understood in this negative sense—as simply fewer total differences—can be defined irrespective of any form or model and without even specifying which differences are reduced. A comparison of specifically delineated differences is not necessary, only the fact of some reduction. It is in this respect merely a quantitative rather than a qualitative determination of similarity, and consequently it lacks the formal and aesthetic aspects of our everyday conception of similarity.

    Comment by Asher Kay — 30 May 2012 @ 10:29 pm

  22. It must really be necessary to enter into Deacon’s lifeworld. That quote probably made total sense when I was reading the book, but now? Here’s another puzzlement: this sense of constraints serving to reduce the number of possible states. The one possible state of primary concern in an entropically-trending universe is the state of total disorder. This state isn’t just possible; it’s inevitable via homeodynamics. Constraints block or partition the system, counteracting its orthograde trajectory with contragrade dynamics. But this sort of partitioning doesn’t reduce the number of states; it increases them. So, e.g., a volume of gas on a gradient of different temperatures will eventually equilibrate at the one temperature, but if you impose a constraint — e.g., put a barrier in the middle of the container — then the system can maintain itself at two different temperatures — two different states rather than one. The propagation of constraints increases the number of partitions resisting orthograde equilibration, which increases the number of states in the overall system. At least that’s my understanding, which doesn’t seem to fit with Deacon’s.

    Comment by ktismatics — 30 May 2012 @ 10:46 pm

  23. Or at least his god subroutine.

    Entropy is totally fucking vexed. Here’s one way to understand it that gives you Deacon’s basic view:

    Imagine that for a box of air, you can record a chart of the position and trajectory of each molecule at any given instant in time. Now imagine a huge pile of such charts, one for each possible combination of positions and trajectories. A high-entropy box of air is one for which the probability of any given chart being the current one is about the same. A low-entropy box is one where a small group of charts are much more likely.

    The counter-intuitive thing is that a box of air with a temperature gradient actually ends up making a bunch of points in the state space (charts) much more unlikely.

    I guess a simpler answer is that it’s misleading to think of a temperature as a state. If you imagine the total energy in the box as constant (maybe kept that way by heating one side and cooling the other), what you see is a change in the distribution of energy. It’s uniform at equilibrium and less “spread out” when the heating and cooling are in effect. Less spread = fewer states.

    Comment by Asher Kay — 30 May 2012 @ 11:38 pm

  24. I forgot to say – the reason it’s misleading is that if temperature is a state of the whole box, it’s the same in both cases. But if it’s a state in some sub-section of the box, then you’re not looking at the whole system.

    Comment by Asher Kay — 30 May 2012 @ 11:40 pm

  25. Yes that’s helpful. Randomness is unlimited freedom of motion; by introducing a constraint into a system, equilibration is inhibited, limiting the freedom of movement of electrons. But the constraint partitions the overall space into two energy gradients — two states of equilibrium rather than one. As such constraint is anti-entropic and pivotal in Deacon’s theory.

    So thinking about mental processes of categorization and comparison and so on… Deacon argues that it’s a mistake to try mapping the mental back to the material, the ideas and perceptions back to the activities of specific groups of neurons, via the usual scientific procedure of decomposing the whole into parts. Instead it’s better to think of neural activity as distributed, nearly random at low energy levels like when you’re asleep. When the situation calls for it — an environmental circumstance or a self-maintaining need — the brain kicks it up a notch. Increased blood flow in the brain pushes it into a far-from-equilibrium condition. This in turn triggers the self-organization of some or all of the neural space into a more constrained and partitioned state. So instead of the visual representation system seeing nothing at all, or seeing the buzzing confusion characteristic of high-entropy randomness, the neural network partitions itself into subsystems that separate out different gradations of luminance, edges, specific objects, categories of objects, comparisons with other objects, etc. It’s always the same congeries of neurons, but with increased energy coming in from the blood substrate they partition themselves into different subsystems that make possible a corresponding partitioning of the world being perceived.

    This doesn’t have to mean that the world isn’t already partitioned without being so perceived by a mind. It just means that the partitioning activity going on in the mind isn’t mapped point-for-point to the information it receives from the environment. As Deacon points out, the mind can engage the world at different levels of intensity, from a general vague awareness of one’s surroundings to a sharp attentiveness to detail. Even within a fully alert state it’s only the center of visual attention that comes into sharp focus; color vision and resolution degrade rapidly outside of the focal area. Deacon’s point would be that there’s no use in looking for the focus neurons, that focus is a diffuse emergent property of a systemic self-organization built on top of lower-order self-organization.

    If I think about it this way then I get Deacon’s constraint idea better. It fits better with neurological evidence of brain activation than does the idea of mapping specific percepts and ideas to specific neural locations, and it accounts better for varying overall levels of mental activity happening in the same brain.

    Comment by ktismatics — 31 May 2012 @ 6:35 am

  26. How is a mind like a hammer? A mind is like a hammer in that it manifests different properties depending on the use to which it’s being put. A mind is different from a hammer in that, while the hammer qua physical object is always the same regardless of whether or not it’s being used to hammer something, the mind’s intrinsic organizational state changes in accord with the telos in which it participates.

    Comment by ktismatics — 31 May 2012 @ 7:36 am

  27. Yes. The key insight, in the terms you use in #25, is that if the world wasn’t partitioned, the mind (in my view, the procedural network of the brain) couldn’t be partitioned. So even if there’s no “pure” correlation (due to vagueness, attention, or just the nature of signal degradation intrinsic to “work”), the partitioning of the mental processes will be determined by the processes generating the perceptual information. You can see in here the “we are the world” answer to Kant that I so favor. Deacon’s addition is that it’s the absences (constraints) that are important to the partitioning.

    If Deacon is right, perhaps we should be more attuned to what’s *not* activated in the brain in the situation being studied.

    #26 – Yeah. That’s a nice way of summarizing how the thermo-morpho-teleo levels work. The “properties”, in essence, are procedural.

    Comment by Asher Kay — 31 May 2012 @ 3:45 pm

  28. I’m not persuaded by the “not activated” proposal, especially now that you’ve convinced me that constraints limit the freedom of random entropy. Old-school point-to-point correspondence between neurons and percepts would emphasize the not-activated status of neurons that map onto features that are not present in the environmental array. The baseline absence would serve as the background against which the relatively few activated neurons stand out to indicate what is present. This is the position that I think Deacon is arguing against. For him, I believe, the background to the perceptual representation of objects isn’t nothingness but noise. Otherwise we are in accord as we move toward future posts on your sight.

    But now I’ve just received a comment from the universe on this thread. It comes from Downriver, a 1991 novel by Iain Sinclair. This afternoon, driving toward the book’s telos, I plucked the bookmark from page 355 and read this:

    “A few harsh bars of ‘Dixie’ on the claxon of his horn announced the arrival of our captain… This creature, our self-inflicted Ahab, hitched his pants and lurched, bow-legged, toward us.”

    Comment by ktismatics — 31 May 2012 @ 4:57 pm

  29. Wow, Dixie and Ahab on the same page. I’m surprised he wasn’t eating a soufflé.

    Comment by Asher Kay — 31 May 2012 @ 5:11 pm

  30. I’m not totally persuaded either. Upon reflection, I think the “nothing vs. noise” distinction you’re making ends up being critical, especially in a neural network example. If the network receives a stream of real noise, the “firing rate” or “weight” of the inputs all tend toward the same value. That value (something greater than 0 but not necessarily all that much greater) reflects the probability of firing on any given distinct input, and it creates a uniform background of activation in the network. Nothing is distinct in the outputs, because all inputs have the same firing rate.

    Now introduce a patterned input into the stream. It would seem to be the *increase* in probability of activation at certain inputs (which we’d associate with a “presence”) that allows the network to categorize the stream as patterned and regular. But in this case – if I’m correct – the firing rate of the inputs still associated with “background noise” actually *goes down* when regularity is introduced. Even though the background-connected inputs are the same noise, their overall probability of activation is below what it would be in a pure noise situation because of the introduced regularity. I could be incorrect about this, but the “muting” of the background may be the sort of absence Deacon is talking about — one in which you don’t need to know any properties of the regularity itself.

    The question, as ever, is whether this kind of absence amounts to something meaningfully different than a flip-side of presence.

    Comment by Asher Kay — 1 June 2012 @ 9:23 am

  31. Yes that’s my understanding of how neural net simulations work as well. Through multiple iterations the overall model accumulates multi-factor probabilistic “weights” via a factor-analytic or related algorithm, thus partitioning up the noise into information. This sort of solution doesn’t require that each node be assigned to detecting a specific sort of input from the environment, as is the case in rule-based or object-based systems. The net’s “learning” remains distributed throughout the system. I think this kind of simulation does represent a closer approximation to how real brains work. Presumably Deacon had this model in mind toward the later chapters of his book, and he may have back-propagated the solution technique, based on systemic partitioning, to simpler self-organizing systems.

    Comment by ktismatics — 1 June 2012 @ 12:34 pm

  32. You guys are amazing. I wish I could link this at DV, but it might lead the Avenging Angel back here….

    Comment by Carl — 1 June 2012 @ 1:46 pm

  33. We’re just working out our DV shtick off-Broadway, away from the glare of the footlights and of the critics.

    Comment by ktismatics — 1 June 2012 @ 2:20 pm

  34. Exactly. We’re getting everything straight here in the ktismatic mists, and then I’ll do this awesome DV post complete with neural network simulation graphics and John will make all these totally erudite comments, and people will be all like, “shit, those guys are so not confused about everything”.

    Comment by Asher Kay — 1 June 2012 @ 3:32 pm

  35. Anyhow, returning briefly to the post and Dennett, I don’t believe that an entirely pragmatic ontology is adequate. Harman and Bryant have both announced with pride that their ontologies aren’t necessarily “true,” that they should be evaluated according to pragmatic and rhetorical criteria. Do the ideas help us think more creatively? Can we marshall the ideas to achieve critical mass among fellow thinkers, thereby enabling us to promulgate a social construction of truth that will achieve dominance over our intellectual foes? Can we construct different ontologies of the hammer from varying points of view: the carpenter’s, the hammersmith’s, the hardware shopkeeper’s, the physicist’s? I respect the realist agenda that asks: What truths can be asserted about the world that explicitly do not depend for their truth value on the pragmatic interests of the truth-sayers.

    Comment by ktismatics — 3 June 2012 @ 10:19 am

  36. I think people probably work with all the rival theories of truth. They go on what works, what is coherent with what is for them well established truths and at the same time willing to be corrected by confrontation with facts which counter both those ‘truths’. I have just been reading Freud’s essay on Screen Memories. The colours that are too bright, the bread that is too tasty are for him a clear indication that the ‘truth’ is not in the memory. His treatment is fascinating and appalling because in its creative leading it has noveletteish tendencies. ‘You think that is your life but it is just a figment this is what happened’. Freud, you old psychopomp you.

    Comment by ombhurbhuva — 3 June 2012 @ 4:31 pm

  37. I put down my cup and examine my own mind. It is for it to discover the truth. But how? What an abyss of uncertainty whenever the mind feels that some part of it has strayed beyond its own borders; when it, the seeker, is at once the dark region through which it must go seeking, where all its equipment will avail it nothing. Seek? More than that: create. It is face to face with something that does not so far exist, to which it alone can give reality and substance, which it alone can bring into the light of day.

    And I begin again to ask myself what it could have been, this unremembered state which brought with it no logical proof of its existence, but only the sense that it was a happy, that it was a real state in whose presence other states of consciousness melted and vanished. I decide to attempt to make it reappear…

    …and so he takes another little crumb of that too-tasty madeleine…

    Comment by ktismatics — 3 June 2012 @ 8:58 pm

  38. If Alexei were here, he’d probably say that there’s no way to separate truth from normativity.

    I personally have no problem with someone who creates a better model sheerly to have a more accurate way of describing the world. But to evaluate the model, you always have to ask what it can do.

    Comment by Asher Kay — 4 June 2012 @ 2:29 am

  39. I agree that truth is a norm against which statements about the reality of things ought to be evaluated. This too is the position of Brassier: here are the first two numbered paragraphs in his 2011 essay in The Speculative Turn:

    1. The question ‘What is real?’ stands at the crossroads of metaphysics and epistemology. More exactly, it marks the juncture of metaphysics and epistemology with the seal of conceptual representation.

    2. Metaphysics understood as the investigation into what there is intersects with epistemology understood as the enquiry into how we know what there is. This intersection of knowing and being is articulated through a theory of conception that explains how thought gains traction on being.

    What can an ontology do? It can provide true concepts about the reality of the world. Is truth an absolute standard, an ideal? I’d think it can be defined procedurally as criteria for evaluating specific truth statements about reality. In that sense we’re back to pragmatism: ontology as a method for generating true conceptual representations about what is.

    Comment by ktismatics — 4 June 2012 @ 6:29 am

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at WordPress.com.

%d bloggers like this: