Loading ...
Sorry, an error occurred while loading the content.

Re: [ai-philosophy] thought

Expand Messages
  • Glen Sizemore
    ... From: Sergio Navega Subject: Re: [ai-philosophy] thought To: ai-philosophy@yahoogroups.com Date: Saturday, January 26, 2013, 8:08 AM  
    Message 1 of 29 , Jan 27, 2013
    • 0 Attachment


      --- On Sat, 1/26/13, Sergio Navega <snavega@...> wrote:

      From: Sergio Navega <snavega@...>
      Subject: Re: [ai-philosophy] thought
      To: ai-philosophy@yahoogroups.com
      Date: Saturday, January 26, 2013, 8:08 AM

       

      > Anyplace there are concepts and assumptions in a science
      > (i.e., everywhere) philosophy is relevant since concepts and
      > assumptions can only be evaluated via philosophy.
       
      SN: And that's where you show you're wrong. Scientific concepts
      and assumptions are evaluated essentially through experimental
      support, not via philosophy. Of course, they must be logically
      coherent (and this is where philosophy might help), but
      experimental support is a must, and that's what decides if
      the concept is good or just garbage. A case in point (in
      search of empirical corroboration): string theory in Physics.
       

      GS: No, this is where you're wrong. Exactly how is an assumption
      validated by experiment? I think that (b) is the most pertinent
      here, no? 

      5
      a : an assuming that something is true
      b : a fact or statement (as a proposition, axiom, postulate, or notion) 
      taken for granted

      Pretty much says it all, eh, Sergio? 

      > Bennett is an accomplished neuroscientist and radical
      > (Skinnerian) behaviorists have always held that it is
      > neuroscience that will provide a reductionistic treatment of
      > the behavioral phenomena elucidated by behavior analysis
       
      SN: And here is where you show that the worldview of behaviorists
      is composed of black and white only. It's the gray stuff that
      matters! Take a specific behavior. Now take the firing of a
      particular neuron. This is the level of analysis that behaviorists
      are proposing. But all the action happens in the middle of these
      two extremes.


      GS: What in the name of Almighty Yahweh are you talking about? 
      Behaviorism makes no claims about the anatomical units
      appropriate to a physiology of behavior. What you claim is
      absurd.   

       
      SN: And in the middle, what we find are cognitive neuroscientific
      stuff. And this is not "my opinion": it is the way brain
      science is being made today, by 95% of the researchers in
      the world. So what you're saying is that all these guys are
      wrong, that they are doing "nonsense", creating useless concepts.
      Kinda weird position yours is, don't you think?


      GS: Why is it that cognitivists always resort to argumentum ad
      populum? But, no, I don't find my position "weird" except in
      the sense that it is a minority position. 
       
      > First of all, behaviorism is a philosophy - not a science
       
      SN: Finally! At last! Something we agree entirely ;-)


      GS: But so is representationalism (the heart of cognitive
      "science")a philosophy.
       
      > but not quite everyone has his or her head up his or her rectum
       
      SN: From that I infer that you're saying that more than 95% of brain
      researchers in the world are, at the moment, performing colonoscopy
      on themselves (which, let's say, is somewhat useful ;-).
      So the world is filled with misguided guys! How nice that
      behaviorists are still here to save us all!


      GS: Yes, but there is no guarantee that the rescue will be
      successful. Standing in the way is the conceptual error of
      mistaking assumptions for hypotheses or theories. 
       
      > I do know that Dennett is a moron, though.
       
      SN: I wouldn't expect more of you, Glen ;-)


      GS: And you have lived up to all my expectations of you.
       
      > You seem to be making an argument about "unobservables." No?
       
      SN: More than that. When some cognitive scientist talks about
      "short term memory", he/she is creating something that is
      unobservable. But is empirically verifiable! 


      GS: Oh Lord (see, you've forced an atheist to invoke God twice)!
      The only thing that is "verifiable" are the behavioral facts
      said to characterize "STM." It is an assumption that the facts
      are as they are because of some "module" that (miraculously) has
      exactly the character necessary to explain the facts from which
      it is inferred. Amazing! Can you say "dormitive virtue"? As
      Skinner said, "Moliere's audience laughed."    



      SN: And now with
      the help of cognitive neuroscience we're understanding the
      brain areas that execute this fuction, as opposed, for instance
      to long term memory. So this means that something that was
      born as an unobservable concept became empirically corroborable,
      surviving the first treadmill of science, and on it's way to
      be completely explained by reductionist mechanisms. Compare
      that with lots of concepts that were abandoned in the past
      because of lack of empirical support and you get a self-correcting
      way of making science.


      GS: And how, exactly, would the notion that there is a
      specialized "thing" called STM, that is responsible for the data
      collected under the rubric of "investigation of STM," be
      falsified? No matter what data are collected - even if
      surprising - the characteristics of the "STM module" are simply
      altered to reflect the data! It is ironic that under the guise
      of the Popperian philosophy of science, cognitive "scientists"
      adopt a strategy of modifying, post hoc, the concept allegedly
      "tested."    


       
      SN: That's the way science progresses, and here I'm not
      talking about cognitive neuroscience only, but ALL scientific
      disciplines, including Physics, Biology, Astrophysics, etc.


      GS: Your vision of science is laughably sophomoric. 

       
      > and assumptions are not testable - err...because they're
      > assumptions.
       
      SN: That's not the way it works. Assumptions are indirectly subject
      to empirical scrutiny. Take one assumption. From it you derive
      a series of logically sound consequences. You test those
      consequences and they do not match the predictions. Then
      something is wrong: your deductions or your assumptions!
      Change them and see what you've got. That's how science
      proceeeds.


      GS: What part of the definition of "assumption" eludes you? 

       
      > But the cognitivist says
      > that the physiological similarities are "rule induction." But
      > the physiology measured is not "rules" etc. What is measured is
      > action potentials, of synchronized networks, or whatever. But
      > the cognitivist says, "See! I told you there was 'rule
      > induction!'" That is, whatever they find, they interpret it in
      > light of their specious assumptions.
       
      SN: From the idea of "rule induction" cognitive neuroscientists
      start to list the properties that this whole process must possess.
      This is the "filling up with meat" of the skeleton of a concept.
      Each chunk of meat that you put in that concept is something
      that can be empirically tested, and will be rejected if predictions
      fail. And again, this is not specific to cognitive neuroscience,
      this is how all science works. If you doubt this, get a book
      of Organic Chemistry, or Electromagnetism or Geology and see
      how many "conceptual structures" they use.


      GS: I'm not condemning assumptions and concepts. I am saying that
      there are issues of fact and theory and there are issues that are
      purely conceptual. Mainstream psychology has rejected all but the most
      trivial of conceptual analyses.  
       
      > The assumption is that all behavior results from rule-following.
       
      SN: And this is easily falsifiable, given the level of depth that
      we have today in brain analysis, from the recording of microelectrodes
      to fMRI and other techniques. 

      GS: Hardly. Any physiological data are interpreted from within
      the philosophical assumptions that characterize it.
       
       
      SN: Let me give you another example. Few constructs are more "to the
      core" of cognitive science than "short term memory" and "working
      memory". You measure behavioral responses and sees that Miller's
      magical number 7 (plus or minus 2) is a fact. Behaviorists stop
      there. Nothing further is pursued. Cognitive Neuroscientists
      go on, proposing working memory store, long term memory,
      the process of chunking, task switching, etc. And what do we have
      now? Many, many studies relating these concepts to lower level
      (empirically sound) phenomena, advancing our understanding of
      the whole thing. One simple example among many:
       
      Phase-dependent neuronal coding of objects in short-term memory
       
      In my way of seeing things (and in 95% of researchers in the world)
      this is pure science, advancing our knowledge.
       
      Meanwhile, behaviorists are resting in their chairs, telling
      us that all this is nonsense...

      Sergio Navega


      GS: The facts are not nonsense. What behaviorists do, however, is
      to examine the histories necessary for certain behavioral
      phenomena to be observed. Cognitive "science," for the most part
      treats behavioral phenomena as ahistorical. For example, when you
      train animals in delayed match-to-sample, the "forgetting curve"
      is a function of training. If you train animals with a zero
      second delay and then test at longer delays, the curve is quite
      steep. If you give them training with gradually increasing
      delays, the curve becomes less steep until it starts dropping
      off. Whatever "working memory" is, it is learned. Behaviorism
      differs from cognitive "science" too in that behaviorists have
      a limited set of observable behavioral processes (i.e., it is
      parsimonious) and it interprets complex phenomena using these
      concepts. Cognitive "science," on the other hand, invents
      unobservable processes to fit its need. Dormitive virtue. Look
      it up.   
       
      With Affection,

      Uncle Glen
       
       
       
       
      Sent: Friday, January 25, 2013 6:10 PM
      Subject: Re: [ai-philosophy] thought
       
       



      --- On Fri, 1/25/13, Sergio Navega <snavega@...> wrote:

      From: Sergio Navega <snavega@...>
      Subject: Re: [ai-philosophy] thought
      To: ai-philosophy@yahoogroups.com
      Date: Friday, January 25, 2013, 8:06 AM

       
      SN: Glen, there's a segment of what you say that I agree with
      you. I too am concerned with some things of cognitive science,
      mainly in the segment where this discipline is close to
      philosophy.
       
       
      GS: Anyplace there are concepts and assumptions in a science
      (i.e., everywhere) philosophy is relevant since concepts and
      assumptions can only be evaluated via philosophy. Thus, what
      you say here makes no sense. Sorry...that happens to you a lot.
       
       
       
      SN:There, at that particular junction, we find
      a lot of garbage and nonsense.
       
       
      GS: So much of cognitive "science" and the fields it has
      corrupted is conceptual crap that the field (and those it has
      corrupted) is not even science.
       
      SN: The other extreme is where we find behaviorists (and Bennett
      and Hacker) who act as if humans had no brains.
       
       
      GS: Bennett is an accomplished neuroscientist and radical
      (Skinnerian) behaviorists have always held that it is
      neuroscience that will provide a reductionistic treatment of
      the behavioral phenomena elucidated by behavior analysis.
      Thus, what you say is misguided nonsense. Sorry...that happens
      to you a lot.
       
       
       
      SB: According
      to them, we have only bodies and behaviors, and that's
      what limits their explanatory pursuit. No wonder
      behaviorism is a forgotten scientific endeavor nowadays.
       
       
      GS: First of all, behaviorism is a philosophy - not a science.
      I have already dealt with behaviorism's position on neuroscience.
      And I think you should check into whether or not behavior
      analysis is actually "forgotten." But, you're right in the sense
      that cognitive "science" has corrupted ALMOST everyone whose
      concern involves behavior...but not quite everyone has his or her
      head up his or her rectum.
       
       
      SN: The Mereological fallacy, if applied rigorously to any scientific
      discipline, would undermine most current science.
       
      GS: How so?
       
       
       
      SN: Bennett and Hacker's
      argument can be applied to Physics, Chemistry, Biology,
      Astrophysics and more.
       
      GS: How so?
       
      SN: No such disciplines would be allowed to
      build conceptual (categorical) constructions, even if these
      constructions were empirically justifiable. It would stall
      scientific progress entirely.
       
       
      GS: I'm not convinced you understand the mereological fallacy.
      What you say above seems to make no sense.
       
       
      SN: That's why Bennett and Hacker's
      thoughts are nonsense (as Dennett and Searle pointed out).
       
      GS: I can't say if their arguments are cogent, since you
      have not described them. But, as I said, nothing you have
      said convinces me that you understand the issues. I can't
      speak to Dennett's and Searle's position. I do know that
      Dennett is a moron, though.
       
       
      SN: And in the same way that Physics do today, so does modern cognitive
      neuroscience.
       
       
      GS: You seem to be making an argument about "unobservables." No?
      But the mereological fallacy involves much more than that. It
      involves talking about parts as if they had the properties of
      the whole. It would be like saying that the entropy of a gas
      increases because the entropy of individual molecules is
      increasing. I don't think physics does that. I await further
      exposition on your part.
       
       
      SN: They are building conceptual structures (memory, 
      rule induction, workspace memory, task switching being some examples)
      that are being successfully supported by lower level explanations
      (mainly the binding and/or synchronization of populations
      of neurons).
       
       
      GS: The concepts underlying a science constitute, in part, the
      assumptions of that science (or "science" as the case may be)
      and assumptions are not testable - err...because they're
      assumptions. Take "rule induction." The observation is that,
      for example, an animal (human or otherwise) responds to a novel
      instance of some stimulus as it did to the exemplars by which it
      was trained. Observations of physiology may reveal certain
      similarities between the effects of training exemplars and novel
      exemplars. That is where the facts end. But the cognitivist says
      that the physiological similarities are "rule induction." But
      the physiology measured is not "rules" etc. What is measured is
      action potentials, of synchronized networks, or whatever. But
      the cognitivist says, "See! I told you there was 'rule
      induction!'" That is, whatever they find, they interpret it in
      light of their specious assumptions. 
       
      SN:Take for instance this article, which is currently being published
      in Trends in Cognitive Sciences:
       
       
      This paper relates rule selection (a "cognitive concept") with
      neural synchronization in the prefrontal cortex (a directly
      measurable empirical evidence).
       
       
      GS: But this is not "evidence." There IS no evidence relevant
      to assumptions for assumptions are assumptions. The assumption
      is that all behavior results from rule-following. When some
      regularity is found it is INTERPRETED in terms of the
      assumptions.
       
      SN: This is exactly the very same
      process used in the majority of the hard sciences today (as I said,
      including Physics, Biology, Cosmology, Chemistry, Geology, etc.).
      But not behaviorism. What went wrong? I wonder if it is some
      sort of religious belief, the church of Skinner...
       
      Sergio Navega
       
       
      GS: It is NOT the process used in physics etc. It is the
      assumption that all behavior is "rule-following" that smacks
      of religion. Take the Higgs Boson, for example. It is not an
      assumption, it is an hypothesis that is generated by
      a mathematical theory. The theory predicts that there should
      be a particle that has certain properties. Observation of a
      particle that has these properties is, indeed, a kind of
      evidence. There is a big difference between assumptions and
      hypotheses/theories. Cognitive science muddles this distinction
      and is, thus, junk. Hope this helps.
       
      With magnanimity,
       
      Glen
       
       
       
       
       
       
       
      Sent: Thursday, January 24, 2013 11:48 AM
      Subject: Re: [ai-philosophy] thought
       
       



      --- On Wed, 1/23/13, Sergio Navega mailto:snavega%40gmail.com> wrote:

      From: Sergio Navega mailto:snavega%40gmail.com>
      Subject: Re: [ai-philosophy] thought
      To: mailto:ai-philosophy%40yahoogroups.com
      Date: Wednesday, January 23, 2013, 10:03 AM

       

      SN: Glen, your mentioning of "homunculi" attest that you're
      criticizing something that has been abandoned decades ago.
      GS: The homunculi were always implied - therefore, there was really nothing to EXPLICITLY abandon. Nobody ever EXPLICITLY endorsed homunculi inside the head driving the rest of the body around like a car. And the homunculi are STILL implied. Another name for this conceptual nonsense is the "mereological fallacy" (cf, Bennett and Hacker).

      SN: I'm talking about modern cognitive neuroscience, where theories
      of memory, language, perception, reasoning, categorization,
      rule-induction, etc., are supported by imaging (fMRI, PET, etc.)
      and also by data collection of neural activity (including single
      and multi-neuron electrodes).

      GS: Since the theories are composed of (literally) nonsensical concepts (cf, Bennett and Hacker), you'll excuse me if I'm unimpressed.

      SN: Today's models of the workings of
      populations of neurons are capable of explaining many cognitive
      phenomena.

      GS: Needless to say, I disagree with you. Most of the questions asked by cognitive neuroscience are literal nonsense (cf, Bennett and Hacker). Some of the actual facts are, perhaps, interesting (after divested of the ridiculous assumptions embodied by the concepts), but cognitive neuroscience explains nothing (since its underlying concepts are literal nonsense).
       
      Cordially,
      Glen
       
       

       

      From: Glen Sizemore
      Sent: Tuesday, January 22, 2013 9:40 PM
      To: mailto:ai-philosophy%40yahoogroups.com

      Subject: Re: [ai-philosophy] thought
       
       

      --- On Mon, 1/21/13, Sergio
      Navega mailto:snavega%40gmail.com> wrote:

      From: Sergio
      Navega mailto:snavega%40gmail.com>
      Subject: Re: [ai-philosophy]
      thought
      To: mailto:ai-philosophy%40yahoogroups.com
      Date: Monday,
      January 21, 2013, 2:04 PM

       






      SN: Boy, what a surprise to find 2 viewpoints that I disagree
      so much.
       
      GS: Hardly surprising to me since I've known you for many
      years, and you were always duped by cognitive "science."
       
       
      SN:Of course the
      "hardware" of the brain is important (and I agree that
      the hardware is mostly determined by genetic
      constraints).
       
       
      GS: Ah...a preformationist to boot!
       
       
      SN: But let's
      take this analogy one step beyond: it is the "software" that
      makes the
      thing "behave" the way they do. And by software one has to
      concede
      that the fundamental thing are "experiences and
      interactions".
       
       
      GS: And what, precisely, are reinforcement contingencies?
      They determine much of what an animal experiences and they are, by
      definition, interactions between organism and environment. You
      must be one of the many who like to bad-mouth behaviorism but know
      next to nothing about it. Lot of that going around.
       
       
      SN:Which leads us to
      finally arrive at the important neuroscience of
      cognition, which not only has extraordinary explanatory
      success,
      but also has important multidisciplinary consequences.
       
      GS: Too bad it's a conceptual cesspool. Anyway, though, to
      the hopelessly simple, made-up stuff often seems to explain a
      lot.
       
       
       
      SN: All is
      not lost for behaviorism though: it makes a damn good
      psychological
      therapy, with much greater efficacy than the nutcracks of
      the
      psychoanalysts.
       
       
      GS: Yeah...interesting how natural sciences often lead to
      technologies isn't it? But cognitive "science" could lead to a
      technology of behavior - if there was anything to the implied
      homunculi of cognitive "science." Or, put it this way...cognitive
      science might lead to a technology if mercury atoms were "silvery"
      and "slippery." For if that were true, it might also be true that
      there are little homunculi in the brain that intend, and decide
      and read maps and think etc. etc. etc. etc.
       
       
       
       
       
       


       

      From: Glen
      Sizemore
      Sent: Monday, January 21, 2013 4:11 PM
      To: mailto:ai-philosophy%40yahoogroups.com
      Subject: Re: [ai-philosophy] thought
       
       

      It may be true that
      "'Tis a gift to be simple." However, 'tis no gift to
      be simplistic." And that is exactly how I would
      describe your posts that I have been reading now for,
      what, 5 years...10 years? On the other hand, I think
      that there is merit to what you say. But a fleshed-out
      position would integrate reinforcement principles
      which is exactly what you gloss over. As to language
      acquisition, I think that there is no question that
      heard utterances lead to parts of the utterances being
      a strange kind of conditioned reinforcer. When the
      child utters, for example, phonemes that it has heard,
      the hearing of these utterances automatically
      reinforce the behavior that leads to their production.
      This all requires a bunch of neuroanatomical
      structures like the basal ganglia etc. The segment of
      neurobiology concerned with analyzing the
      physiological mediation of operant behavior is an area
      of neuroscience that is relatively uncontaminated by
      the conceptual nonsense that plagues much of
      neuroscience. Of course, many workers in this area
      seem to struggle to obscure the connection to the vast
      literature that constitutes the natural science of
      behavior.

      --- On Sun, 1/20/13, scanlonray
      mailto:rscan%40nycap.rr.com> wrote:

      From:
      scanlonray mailto:rscan%40nycap.rr.com>
      Subject:
      [ai-philosophy] thought
      To:
      mailto:ai-philosophy%40yahoogroups.com
      Date: Sunday,
      January 20, 2013, 11:11 AM

       

      The apparatus for producing speech is constructed
      by the genome More than one hundred pairs of
      orofacial, laryngeal, pharyngeal, and respiratory
      muscles together with all the neural groups that
      drive them.

      A motor neuron arrives at
      striated muscle fiber and terminates on it. When an
      axonal pulse arrives at the terminal the muscle
      fiber contracts. It is apparent that we can separate
      tahe motor neurons from the muscle fibers and view
      them abstractly as a set of terminals connected to a
      computer, and activate the muscles by a computer
      program, a program of
      incredible complexity, but
      not beyond human abilities. This is a thing that can
      be written by a computer programmer as a part of the
      days work..

      Any objection about the
      complexity of the neural network involved is
      answered by a reference to the flight of the brush
      turkey.

      We proceed under the assumption that
      the genome can construct the required circuitry and
      its flexibility.

      At this point we introduce
      the reticular thalamic nucleus, a thin layer of
      neurons that surrounds the thalamus like a blanket.
      Every motor program, every sensory group must enter
      the thalamus through this blanket. The neurons of
      the blanket are all alike in the their output is
      solely inhibitory.

      Some spoken utterance
      enters the thalamus. At the same time a snippet of
      baby talk enters. If there is any resemblance, the
      equivalence is recorded. If some agreement is
      reached the baby talk is allowed to proceed. Day by
      day the resemblances are recorded. It seems that
      nothing is gained, but link by link ground is
      gained. A language is learned.
      This continues for
      life

      When we come up against speakers with a
      different accent we adopt or reject slight
      differences always striking toward a weighted
      statistical average. The thalamus continues to
      examine each motor program as if passes through when
      we attempt an utterance that violates our maintained
      average of what is "right", the thalamic reticular
      nucleus is activated, and the utterance vanishes
      into the great cloud of things left
      unsaid..

    • Glen Sizemore
      Read some Oyama. Developmental Systems Theory.  ... From: Sergio Navega Subject: Re: [ai-philosophy] thought To:
      Message 2 of 29 , Jan 27, 2013
      • 0 Attachment
        Read some Oyama. Developmental Systems Theory. 

        --- On Sat, 1/26/13, Sergio Navega <snavega@...> wrote:

        From: Sergio Navega <snavega@...>
        Subject: Re: [ai-philosophy] thought
        To: ai-philosophy@yahoogroups.com
        Date: Saturday, January 26, 2013, 8:12 AM

         

        Uncle Glen, isn't it time for your afternoon tea?
        And don't forget to take your medications, ;-)
         
        Sergio Navega
         
         
         
         
        Sent: Friday, January 25, 2013 7:13 PM
        Subject: Re: [ai-philosophy] thought
         
         

        Both Navega and Scanlon are preformationists but, at least, Navega pays lip-service to the environment. Scanlon doesn't even pay lip-service. His position is laughably naive. Not that Navega's position isn't worth a chuckle or two. For example, if you excise a forelimb bud and transplant it to the hind limp, you get a hind limb. The expression of the genes depends on the local embryological environment. Genes are the same, no? But then, the preformationist says, "Well...yeah...the rules are in the genes but the environment allows the expression of THE rule." Oy Vey! Can you get anymore imbued with "belief"?  Maybe it's "evolution's plan," eh, Ray? But how could you tell? Maybe it's God's plan? But maybe there is no plan. Maybe development at all biological levels involves genes as merely ONE variable in an exceedingly complex dynamic developmental process; sometimes really important (my adopted son looks like a dark-skinned Romanian while my wife and I are blond), sometimes much less important (he speaks English pretty well, like my wife and me, but his Romanian sucks - like my wife's and mine). But then, you can always claim (especially when you refuse to say very much aside from pseudo-profound jabber) that all these plans are in the genes. Again, talk about "beliefs"! Ray is an amusing example of a not-so-well-informed hypocrite. I'm not sure which aspect of his "intellectual" endeavors is more troublesome.  
         
         
        Cordially,
         
        Uncle Glen
         


        --- On Fri, 1/25/13, scanlonray <rscan@...> wrote:

        From: scanlonray <rscan@...>
        Subject: [ai-philosophy] thought
        To: ai-philosophy@yahoogroups.com
        Date: Friday, January 25, 2013, 12:48 PM

         

        But here's the thing, Ray: neurons connect to each other
        according to genetic rules, but the expression of these
        rules depend heavily on environmental factors. Stimuli
        received by an animal is what drives the need to create
        or reinforce a particular axon-dendrite connection.
        And this is important: it is the genome that says how this
        creation/reinforcement is made, but it was the stimuli from
        the environment that required it. Therefore brains are a
        function of genomic and environmental interaction.
        Sergio Navega

        This is a straightforward statement of a position that I understand but do no share.
        If you believe in the tabula erasa then all else follows. If instead you are driven by the present day knowledge of the genome than a different scenario presents itself. For this we turn to the brush turkey, A living creature that enters the world with no help from the environment, with nothing but a brain constructed by the genome, a creature that does rather well to meet the environment.

        "Therefore brains are a function of genomic and aenvironmentsl interaction" This is a statement of belief, which is not present in the "brush turkey". What are we to say about `imprinting", when the brush turkey does not imprint, flies without instruction? A bird that eats without instruction? What are we to say about "reinforcement" to a bird that gets none?

      • Sergio Navega
        Glen, you must notice that assumptions and conceptual structures in science in general are eventually subject to rejection and substitution when competing
        Message 3 of 29 , Jan 28, 2013
        • 0 Attachment
          Glen, you must notice that assumptions and conceptual
          structures in science in general are eventually subject to
          rejection and substitution when competing theories appear and
          turns them into outdated material. This has been so since the
          substitution of Ptolemaic by Copernican models of the solar
          system. And cognitive neuroscience is doing just that, and
          today more than ever, because of the strong link with studies
          of the dynamics of populations of neurons. Take a look at
          the table of contents of a first-class journal like "Trends
          in Cognitive Sciences": you'll find lots of articles with
          neuroscience stuff. So the skinnerian "philosophy of behavior
          analysis" is now a topic in the history of science, in the
          same chapter about alchemy ;-)
           
          Sergio Navega
           
           
           
          Sent: Sunday, January 27, 2013 3:16 PM
          Subject: Re: [ai-philosophy] thought
           
           



          --- On Sat, 1/26/13, Sergio Navega <snavega@...> wrote:

          From: Sergio Navega <snavega@...>
          Subject: Re: [ai-philosophy] thought
          To: ai-philosophy@yahoogroups.com
          Date: Saturday, January 26, 2013, 8:08 AM

           
          > Anyplace there are concepts and assumptions in a science
          > (i.e., everywhere) philosophy is relevant since concepts and
          > assumptions can only be evaluated via philosophy.
           
          SN: And that's where you show you're wrong. Scientific concepts
          and assumptions are evaluated essentially through experimental
          support, not via philosophy. Of course, they must be logically
          coherent (and this is where philosophy might help), but
          experimental support is a must, and that's what decides if
          the concept is good or just garbage. A case in point (in
          search of empirical corroboration): string theory in Physics.
           
           
          GS: No, this is where you're wrong. Exactly how is an assumption
          validated by experiment? I think that (b) is the most pertinent
          here, no?
           
          5
          a : an assuming that something is true
          b : a fact or statement (as a proposition, axiom, postulate, or notion)
          taken for granted
           
          Pretty much says it all, eh, Sergio?
           
          > Bennett is an accomplished neuroscientist and radical
          > (Skinnerian) behaviorists have always held that it is
          > neuroscience that will provide a reductionistic treatment of
          > the behavioral phenomena elucidated by behavior analysis
           
          SN: And here is where you show that the worldview of behaviorists
          is composed of black and white only. It's the gray stuff that
          matters! Take a specific behavior. Now take the firing of a
          particular neuron. This is the level of analysis that behaviorists
          are proposing. But all the action happens in the middle of these
          two extremes.
           
           
          GS: What in the name of Almighty Yahweh are you talking about?
          Behaviorism makes no claims about the anatomical units
          appropriate to a physiology of behavior. What you claim is
          absurd.  
           
           
          SN: And in the middle, what we find are cognitive neuroscientific
          stuff. And this is not "my opinion": it is the way brain
          science is being made today, by 95% of the researchers in
          the world. So what you're saying is that all these guys are
          wrong, that they are doing "nonsense", creating useless concepts.
          Kinda weird position yours is, don't you think?
           
           
          GS: Why is it that cognitivists always resort to argumentum ad
          populum? But, no, I don't find my position "weird" except in
          the sense that it is a minority position.
           
          > First of all, behaviorism is a philosophy - not a science
           
          SN: Finally! At last! Something we agree entirely ;-)
           
           
          GS: But so is representationalism (the heart of cognitive
          "science")a philosophy.
           
          > but not quite everyone has his or her head up his or her rectum
           
          SN: From that I infer that you're saying that more than 95% of brain
          researchers in the world are, at the moment, performing colonoscopy
          on themselves (which, let's say, is somewhat useful ;-).
          So the world is filled with misguided guys! How nice that
          behaviorists are still here to save us all!
           
           
          GS: Yes, but there is no guarantee that the rescue will be
          successful. Standing in the way is the conceptual error of
          mistaking assumptions for hypotheses or theories.
           
          > I do know that Dennett is a moron, though.
           
          SN: I wouldn't expect more of you, Glen ;-)
           
           
          GS: And you have lived up to all my expectations of you.
           
          > You seem to be making an argument about "unobservables." No?
           
          SN: More than that. When some cognitive scientist talks about
          "short term memory", he/she is creating something that is
          unobservable. But is empirically verifiable!
           
           
          GS: Oh Lord (see, you've forced an atheist to invoke God twice)!
          The only thing that is "verifiable" are the behavioral facts
          said to characterize "STM." It is an assumption that the facts
          are as they are because of some "module" that (miraculously) has
          exactly the character necessary to explain the facts from which
          it is inferred. Amazing! Can you say "dormitive virtue"? As
          Skinner said, "Moliere's audience laughed."   
           
           
           
          SN: And now with
          the help of cognitive neuroscience we're understanding the
          brain areas that execute this fuction, as opposed, for instance
          to long term memory. So this means that something that was
          born as an unobservable concept became empirically corroborable,
          surviving the first treadmill of science, and on it's way to
          be completely explained by reductionist mechanisms. Compare
          that with lots of concepts that were abandoned in the past
          because of lack of empirical support and you get a self-correcting
          way of making science.
           
           
          GS: And how, exactly, would the notion that there is a
          specialized "thing" called STM, that is responsible for the data
          collected under the rubric of "investigation of STM," be
          falsified? No matter what data are collected - even if
          surprising - the characteristics of the "STM module" are simply
          altered to reflect the data! It is ironic that under the guise
          of the Popperian philosophy of science, cognitive "scientists"
          adopt a strategy of modifying, post hoc, the concept allegedly
          "tested."   
           
           
           
          SN: That's the way science progresses, and here I'm not
          talking about cognitive neuroscience only, but ALL scientific
          disciplines, including Physics, Biology, Astrophysics, etc.
           
           
          GS: Your vision of science is laughably sophomoric.
           
           
          > and assumptions are not testable - err...because they're
          > assumptions.
           
          SN: That's not the way it works. Assumptions are indirectly subject
          to empirical scrutiny. Take one assumption. From it you derive
          a series of logically sound consequences. You test those
          consequences and they do not match the predictions. Then
          something is wrong: your deductions or your assumptions!
          Change them and see what you've got. That's how science
          proceeeds.
           
           
          GS: What part of the definition of "assumption" eludes you?
           
           
          > But the cognitivist says
          > that the physiological similarities are "rule induction." But
          > the physiology measured is not "rules" etc. What is measured is
          > action potentials, of synchronized networks, or whatever. But
          > the cognitivist says, "See! I told you there was 'rule
          > induction!'" That is, whatever they find, they interpret it in
          > light of their specious assumptions.
           
          SN: From the idea of "rule induction" cognitive neuroscientists
          start to list the properties that this whole process must possess.
          This is the "filling up with meat" of the skeleton of a concept.
          Each chunk of meat that you put in that concept is something
          that can be empirically tested, and will be rejected if predictions
          fail. And again, this is not specific to cognitive neuroscience,
          this is how all science works. If you doubt this, get a book
          of Organic Chemistry, or Electromagnetism or Geology and see
          how many "conceptual structures" they use.
           
           
          GS: I'm not condemning assumptions and concepts. I am saying that
          there are issues of fact and theory and there are issues that are
          purely conceptual. Mainstream psychology has rejected all but the most
          trivial of conceptual analyses. 
           
          > The assumption is that all behavior results from rule-following.
           
          SN: And this is easily falsifiable, given the level of depth that
          we have today in brain analysis, from the recording of microelectrodes
          to fMRI and other techniques. 
           
          GS: Hardly. Any physiological data are interpreted from within
          the philosophical assumptions that characterize it.
           
           
          SN: Let me give you another example. Few constructs are more "to the
          core" of cognitive science than "short term memory" and "working
          memory". You measure behavioral responses and sees that Miller's
          magical number 7 (plus or minus 2) is a fact. Behaviorists stop
          there. Nothing further is pursued. Cognitive Neuroscientists
          go on, proposing working memory store, long term memory,
          the process of chunking, task switching, etc. And what do we have
          now? Many, many studies relating these concepts to lower level
          (empirically sound) phenomena, advancing our understanding of
          the whole thing. One simple example among many:
           
          Phase-dependent neuronal coding of objects in short-term memory
           
          In my way of seeing things (and in 95% of researchers in the world)
          this is pure science, advancing our knowledge.
           
          Meanwhile, behaviorists are resting in their chairs, telling
          us that all this is nonsense...
           
          Sergio Navega
           
           
          GS: The facts are not nonsense. What behaviorists do, however, is
          to examine the histories necessary for certain behavioral
          phenomena to be observed. Cognitive "science," for the most part
          treats behavioral phenomena as ahistorical. For example, when you
          train animals in delayed match-to-sample, the "forgetting curve"
          is a function of training. If you train animals with a zero
          second delay and then test at longer delays, the curve is quite
          steep. If you give them training with gradually increasing
          delays, the curve becomes less steep until it starts dropping
          off. Whatever "working memory" is, it is learned. Behaviorism
          differs from cognitive "science" too in that behaviorists have
          a limited set of observable behavioral processes (i.e., it is
          parsimonious) and it interprets complex phenomena using these
          concepts. Cognitive "science," on the other hand, invents
          unobservable processes to fit its need. Dormitive virtue. Look
          it up.  
           
          With Affection,
           
          Uncle Glen
           
           
           
           
          Sent: Friday, January 25, 2013 6:10 PM
          Subject: Re: [ai-philosophy] thought
           
           



          --- On Fri, 1/25/13, Sergio Navega <snavega@...> wrote:

          From: Sergio Navega <snavega@...>
          Subject: Re: [ai-philosophy] thought
          To: ai-philosophy@yahoogroups.com
          Date: Friday, January 25, 2013, 8:06 AM

           
          SN: Glen, there's a segment of what you say that I agree with
          you. I too am concerned with some things of cognitive science,
          mainly in the segment where this discipline is close to
          philosophy.
           
           
          GS: Anyplace there are concepts and assumptions in a science
          (i.e., everywhere) philosophy is relevant since concepts and
          assumptions can only be evaluated via philosophy. Thus, what
          you say here makes no sense. Sorry...that happens to you a lot.
           
           
           
          SN:There, at that particular junction, we find
          a lot of garbage and nonsense.
           
           
          GS: So much of cognitive "science" and the fields it has
          corrupted is conceptual crap that the field (and those it has
          corrupted) is not even science.
           
          SN: The other extreme is where we find behaviorists (and Bennett
          and Hacker) who act as if humans had no brains.
           
           
          GS: Bennett is an accomplished neuroscientist and radical
          (Skinnerian) behaviorists have always held that it is
          neuroscience that will provide a reductionistic treatment of
          the behavioral phenomena elucidated by behavior analysis.
          Thus, what you say is misguided nonsense. Sorry...that happens
          to you a lot.
           
           
           
          SB: According
          to them, we have only bodies and behaviors, and that's
          what limits their explanatory pursuit. No wonder
          behaviorism is a forgotten scientific endeavor nowadays.
           
           
          GS: First of all, behaviorism is a philosophy - not a science.
          I have already dealt with behaviorism's position on neuroscience.
          And I think you should check into whether or not behavior
          analysis is actually "forgotten." But, you're right in the sense
          that cognitive "science" has corrupted ALMOST everyone whose
          concern involves behavior...but not quite everyone has his or her
          head up his or her rectum.
           
           
          SN: The Mereological fallacy, if applied rigorously to any scientific
          discipline, would undermine most current science.
           
          GS: How so?
           
           
           
          SN: Bennett and Hacker's
          argument can be applied to Physics, Chemistry, Biology,
          Astrophysics and more.
           
          GS: How so?
           
          SN: No such disciplines would be allowed to
          build conceptual (categorical) constructions, even if these
          constructions were empirically justifiable. It would stall
          scientific progress entirely.
           
           
          GS: I'm not convinced you understand the mereological fallacy.
          What you say above seems to make no sense.
           
           
          SN: That's why Bennett and Hacker's
          thoughts are nonsense (as Dennett and Searle pointed out).
           
          GS: I can't say if their arguments are cogent, since you
          have not described them. But, as I said, nothing you have
          said convinces me that you understand the issues. I can't
          speak to Dennett's and Searle's position. I do know that
          Dennett is a moron, though.
           
           
          SN: And in the same way that Physics do today, so does modern cognitive
          neuroscience.
           
           
          GS: You seem to be making an argument about "unobservables." No?
          But the mereological fallacy involves much more than that. It
          involves talking about parts as if they had the properties of
          the whole. It would be like saying that the entropy of a gas
          increases because the entropy of individual molecules is
          increasing. I don't think physics does that. I await further
          exposition on your part.
           
           
          SN: They are building conceptual structures (memory, 
          rule induction, workspace memory, task switching being some examples)
          that are being successfully supported by lower level explanations
          (mainly the binding and/or synchronization of populations
          of neurons).
           
           
          GS: The concepts underlying a science constitute, in part, the
          assumptions of that science (or "science" as the case may be)
          and assumptions are not testable - err...because they're
          assumptions. Take "rule induction." The observation is that,
          for example, an animal (human or otherwise) responds to a novel
          instance of some stimulus as it did to the exemplars by which it
          was trained. Observations of physiology may reveal certain
          similarities between the effects of training exemplars and novel
          exemplars. That is where the facts end. But the cognitivist says
          that the physiological similarities are "rule induction." But
          the physiology measured is not "rules" etc. What is measured is
          action potentials, of synchronized networks, or whatever. But
          the cognitivist says, "See! I told you there was 'rule
          induction!'" That is, whatever they find, they interpret it in
          light of their specious assumptions. 
           
          SN:Take for instance this article, which is currently being published
          in Trends in Cognitive Sciences:
           
           
          This paper relates rule selection (a "cognitive concept") with
          neural synchronization in the prefrontal cortex (a directly
          measurable empirical evidence).
           
           
          GS: But this is not "evidence." There IS no evidence relevant
          to assumptions for assumptions are assumptions. The assumption
          is that all behavior results from rule-following. When some
          regularity is found it is INTERPRETED in terms of the
          assumptions.
           
          SN: This is exactly the very same
          process used in the majority of the hard sciences today (as I said,
          including Physics, Biology, Cosmology, Chemistry, Geology, etc.).
          But not behaviorism. What went wrong? I wonder if it is some
          sort of religious belief, the church of Skinner...
           
          Sergio Navega
           
           
          GS: It is NOT the process used in physics etc. It is the
          assumption that all behavior is "rule-following" that smacks
          of religion. Take the Higgs Boson, for example. It is not an
          assumption, it is an hypothesis that is generated by
          a mathematical theory. The theory predicts that there should
          be a particle that has certain properties. Observation of a
          particle that has these properties is, indeed, a kind of
          evidence. There is a big difference between assumptions and
          hypotheses/theories. Cognitive science muddles this distinction
          and is, thus, junk. Hope this helps.
           
          With magnanimity,
           
          Glen
           
           
           
           
           
           
           
          Sent: Thursday, January 24, 2013 11:48 AM
          Subject: Re: [ai-philosophy] thought
           
           



          --- On Wed, 1/23/13, Sergio Navega mailto:snavega%40gmail.com> wrote:

          From: Sergio Navega mailto:snavega%40gmail.com>
          Subject: Re: [ai-philosophy] thought
          To: mailto:ai-philosophy%40yahoogroups.com
          Date: Wednesday, January 23, 2013, 10:03 AM

           

          SN: Glen, your mentioning of "homunculi" attest that you're
          criticizing something that has been abandoned decades ago.
          GS: The homunculi were always implied - therefore, there was really nothing to EXPLICITLY abandon. Nobody ever EXPLICITLY endorsed homunculi inside the head driving the rest of the body around like a car. And the homunculi are STILL implied. Another name for this conceptual nonsense is the "mereological fallacy" (cf, Bennett and Hacker).

          SN: I'm talking about modern cognitive neuroscience, where theories
          of memory, language, perception, reasoning, categorization,
          rule-induction, etc., are supported by imaging (fMRI, PET, etc.)
          and also by data collection of neural activity (including single
          and multi-neuron electrodes).

          GS: Since the theories are composed of (literally) nonsensical concepts (cf, Bennett and Hacker), you'll excuse me if I'm unimpressed.

          SN: Today's models of the workings of
          populations of neurons are capable of explaining many cognitive
          phenomena.

          GS: Needless to say, I disagree with you. Most of the questions asked by cognitive neuroscience are literal nonsense (cf, Bennett and Hacker). Some of the actual facts are, perhaps, interesting (after divested of the ridiculous assumptions embodied by the concepts), but cognitive neuroscience explains nothing (since its underlying concepts are literal nonsense).
           
          Cordially,
          Glen
           
           

           

          From: Glen Sizemore
          Sent: Tuesday, January 22, 2013 9:40 PM
          To: mailto:ai-philosophy%40yahoogroups.com

          Subject: Re: [ai-philosophy] thought
           
           

          --- On Mon, 1/21/13, Sergio
          Navega mailto:snavega%40gmail.com> wrote:

          From: Sergio
          Navega mailto:snavega%40gmail.com>
          Subject: Re: [ai-philosophy]
          thought
          To: mailto:ai-philosophy%40yahoogroups.com
          Date: Monday,
          January 21, 2013, 2:04 PM

           






          SN: Boy, what a surprise to find 2 viewpoints that I disagree
          so much.
           
          GS: Hardly surprising to me since I've known you for many
          years, and you were always duped by cognitive "science."
           
           
          SN:Of course the
          "hardware" of the brain is important (and I agree that
          the hardware is mostly determined by genetic
          constraints).
           
           
          GS: Ah...a preformationist to boot!
           
           
          SN: But let's
          take this analogy one step beyond: it is the "software" that
          makes the
          thing "behave" the way they do. And by software one has to
          concede
          that the fundamental thing are "experiences and
          interactions".
           
           
          GS: And what, precisely, are reinforcement contingencies?
          They determine much of what an animal experiences and they are, by
          definition, interactions between organism and environment. You
          must be one of the many who like to bad-mouth behaviorism but know
          next to nothing about it. Lot of that going around.
           
           
          SN:Which leads us to
          finally arrive at the important neuroscience of
          cognition, which not only has extraordinary explanatory
          success,
          but also has important multidisciplinary consequences.
           
          GS: Too bad it's a conceptual cesspool. Anyway, though, to
          the hopelessly simple, made-up stuff often seems to explain a
          lot.
           
           
           
          SN: All is
          not lost for behaviorism though: it makes a damn good
          psychological
          therapy, with much greater efficacy than the nutcracks of
          the
          psychoanalysts.
           
           
          GS: Yeah...interesting how natural sciences often lead to
          technologies isn't it? But cognitive "science" could lead to a
          technology of behavior - if there was anything to the implied
          homunculi of cognitive "science." Or, put it this way...cognitive
          science might lead to a technology if mercury atoms were "silvery"
          and "slippery." For if that were true, it might also be true that
          there are little homunculi in the brain that intend, and decide
          and read maps and think etc. etc. etc. etc.
           
           
           
           
           
           


           

          From: Glen
          Sizemore
          Sent: Monday, January 21, 2013 4:11 PM
          To: mailto:ai-philosophy%40yahoogroups.com
          Subject: Re: [ai-philosophy] thought
           
           

          It may be true that
          "'Tis a gift to be simple." However, 'tis no gift to
          be simplistic." And that is exactly how I would
          describe your posts that I have been reading now for,
          what, 5 years...10 years? On the other hand, I think
          that there is merit to what you say. But a fleshed-out
          position would integrate reinforcement principles
          which is exactly what you gloss over. As to language
          acquisition, I think that there is no question that
          heard utterances lead to parts of the utterances being
          a strange kind of conditioned reinforcer. When the
          child utters, for example, phonemes that it has heard,
          the hearing of these utterances automatically
          reinforce the behavior that leads to their production.
          This all requires a bunch of neuroanatomical
          structures like the basal ganglia etc. The segment of
          neurobiology concerned with analyzing the
          physiological mediation of operant behavior is an area
          of neuroscience that is relatively uncontaminated by
          the conceptual nonsense that plagues much of
          neuroscience. Of course, many workers in this area
          seem to struggle to obscure the connection to the vast
          literature that constitutes the natural science of
          behavior.

          --- On Sun, 1/20/13, scanlonray
          mailto:rscan%40nycap.rr.com> wrote:

          From:
          scanlonray mailto:rscan%40nycap.rr.com>
          Subject:
          [ai-philosophy] thought
          To:
          mailto:ai-philosophy%40yahoogroups.com
          Date: Sunday,
          January 20, 2013, 11:11 AM

           

          The apparatus for producing speech is constructed
          by the genome More than one hundred pairs of
          orofacial, laryngeal, pharyngeal, and respiratory
          muscles together with all the neural groups that
          drive them.

          A motor neuron arrives at
          striated muscle fiber and terminates on it. When an
          axonal pulse arrives at the terminal the muscle
          fiber contracts. It is apparent that we can separate
          tahe motor neurons from the muscle fibers and view
          them abstractly as a set of terminals connected to a
          computer, and activate the muscles by a computer
          program,

          (Message over 64 KB, truncated)

        • scanlonray
          I thought that a few references to the genome and the brush turkey of Australia ahould be enough to start up s discussion, but I guess not. This is the century
          Message 4 of 29 , Feb 15, 2013
          • 0 Attachment
            I thought that a few references to the genome and the brush turkey of Australia ahould be enough to start up s discussion, but I guess not. This is the century of the genome. And we are given a brain driven organism, completely constructed by the genome so that it is ready to deal with the universe.

            Glen, before you slap on your three-cornered hat and spring to the defense of behaviorism, give a thought to your position as the last leaf upon the tree.
          Your message has been successfully submitted and would be delivered to recipients shortly.