90.9 WBUR - Boston's NPR news station
Top Stories:
The Brain As An Analogy Machine

How humans think.  The human brain as an analogy machine.

Brain Art showcases prizewinners in the 2012 Brain-Art Competition that honors outstanding visualizations of brain research data. The works are by John Van Horn (US), Neda Jahanshad (US), Betty Lee (US), Daniel Margulies (US) and Alexander Schäfer (DE). (Flickr/Ars Electronica)

Brain Art showcases prizewinners in the 2012 Brain-Art Competition that honors outstanding visualizations of brain research data. The works are by John Van Horn (US), Neda Jahanshad (US), Betty Lee (US), Daniel Margulies (US) and Alexander Schäfer (DE). (Flickr/Ars Electronica)

How do we think?  How do our brains make sense of the world?  Of the endless, swirling, changing flow of scenes and situations we confront?

For decades now, while the world has looked at neuro-chemistry and MRIs and synapses flashing, big thinker Douglas Hofstadter has looked for patterns.  With Emmanuel Sander he now says it all comes down to this:  analogy.  Our brains as mighty analogy machines.  Endlessly linking this to that and that to the other.  Analogy as the fundamental fuel and fire of thought.

This hour, On Point:  considering the very core of human thought.

- Tom Ashbrook


Douglas Hofstadter, Distinguished College of Arts and Sciences Professor of Cognitive Science and Comparative Literature at Indiana University. Author of the Pulitzer-Prize-winning book “Gödel, Escher, Bach.” Co-author of “Surfaces and Essences: Analogy as the Fuel and Fire of Thinking.”

Emmanuel Sander, professor of Cognitive and Developmental Psychology at the University of Paris (Saint-Denis), specializing in the study of analogy-making and categorization and their connections to education.  Co-author of “Surfaces and Essences: Analogy as the Fuel and Fire of Thinking.”

From Tom’s Reading List

The New Yorker: Can Super Mario Save Artificial Intelligence? — “Douglas Hoftstadter and Emmanuel Sander implicitly remind us in their new book, ‘Surfaces and Essences,’ that one strategy toward a flexible-learning A.I. might begin by looking at the human gift for analogy. Although Hofstadter and Sander overstate their case (not everything depends on analogy) and never provide an explicit algorithm, it does seem plausible that analogy is one of the most powerful tools that a ten-year-old has in his (or her) mental toolkit for learning new tasks, like a new sport or a new video game.”

Nature: Cognitive Science: Mind as Mirror — “Why (science-fiction writers take note) would we invent new categories and labels for things when we can aid comprehension by borrowing old ones, even if the physical resemblance is negligible? What cognitive scientists Douglas Hofstadter and Emmanuel Sander set out to show is that this sort of elision is not merely a convenience: all our thinking depends on it, from the half-truths of everyday speech (“that always happens to me too!”) to the most abstruse of mathematical reasoning. I was convinced, and the ramifications are often thought-provoking. But when authors tell you the same thing, over and over again, for 500 pages, perhaps you’ll believe it whether it is true or not.”

Excerpt: “Surfaces And Essences”


A video tribute to the ideas of Douglas Hofstadter

THE MIRRORING MIND – by @JasonSilva from Jason Silva on Vimeo.


Please follow our community rules when engaging in comment discussion on this site.
  • Unterthurn

    Each human being is different and to putting everyone, once again, into a cookie cutter style of learning sounds scary. There are so many styles of acquiring information and learning, I hope this is only seeing as one point of view and not a solution for the masses. Our school systems are already one sided enough as it is. Plus the example displayed on the OnPoint page shows that environment and experience make everyone an unique individual. Perhaps the authors published to early and need more research including people of multi-language upbringing and how language is acquired and the understanding of idiomatic expressions.

  • Wm_James_from_Missouri

    Neural networks using weighted connections act a sieves or “loose” filters, allowing objects and concepts to be identified as “good enough” or “close enough” to the original training set and are thereby christened “ acceptable”. Analogy is implicit in the very structure of the Human brain. Our dictionaries define words with phrases, such as, “more at”, or with concepts that hint at “resembling”, leaving the reader to paint their own pictures based on their life’s experience. It is quite amazing, that given the vast differences in personal experiences that we consistently find analogies so d**n useful.

  • Yar

    I am left handed self described as slightly autistic who is a visual thinker. The voice in my mind is a video which is narrated in my own voice. The narration is synthesized from images. Language is not the source of my thoughts.  What gets lost in translation are names of people and places, I have images, sometimes stored in little pieces, for example a face may be stored as a nose, eyes in a different part of the mind, lips, or that little curl of hair on the nap of the neck, all different images. I see some version of the big picture, my comments on this blog over the past two years reflect this world view.  Honeybee democracy, a book on how bee swarms decide on a new hive, dances and stop dances, they mimic individual neurons, insights on how we process information. It is a privileged to have time to think.  Why so many waste time being lulled to sleep by TV is beyond me.  I like radio because the images are better.

    • PithHelmut

      Autism may turn out to be a phenomenon that could give humans additional insights that “conventional” thinking misses, by it being tethered to analogy. Autism may allow a new creation of thought processing, one that may help us adapt to the very new concoctions that we are haphazardly cultivating not only by environmental degradation but also by new paradigms that could flip traditional thought on its head. 

  • ToyYoda

    The brain is an analogy machine?  I’m skeptical.  This would mean that if there is a set of knowledge that has no analogy, then it will never be discovered.  Ask yourself if this is true.  You will need to prove this.  Or you can find counter examples to show that it is false.  

    1. Complex numbers were not  born out of analogies but necessity to solve  polynomial equations.  The “analogous” geometric interpretation came later.

    2. The “invention” of non-euclidean geometry did not come by the analogy of space bending.  It came by looking at the logical consequence of modify ing Euclidean parallel postulate.  

    In fact, the mind as an analogy machine seems suspiciously like Kant’s “mistake” that there could be no geometry other than Euclidean because are brains are built so that it cannot conceive of anything else.  (Mistake is in quotes because I don’t believe Kant actually said that if you read the source text.)

    3.  Finally, How does a baby learn when he/she has no prior analogies?  Does he come with pre-built analogies? If so, I would hardly call pre-built analogies “analogies” in the normal sense of the word, but rather instinct or intuition.

    • http://twitter.com/Dragonsong73 Eric R. Duncan

      As you are talking about the interplay of neurons and various hormones in various amounts and attributing to them terms such as love faith and happiness using a process of symbolic manipulations either written or vocal analogs to one another. Then yes it’s an analogy machine. Pattern recognition is a process of analogization.

      Really what you are getting at is more well covered by Saussure and his theories of language which straight up acknowledges there is no guarantee; because of the vagaries of symbolic expressions like language being transmitted by intent but filtered through the actors experience then re-filtered through the listeners experience and then interpreted/understood that your “Blue” and my “Blue” will ever be the same

      Edit to add: math is also a symbolic manipulation language an analog or metaphor of phenomena of both the micro and macro scales

    • Jasoturner

      I don’t thing these guys would claim analogy is the only way that we reason.  But I think they will contend that analogy is the “fuel and fire” of our ability to reason.

      As for babies, I suspect they see that one parent is like another parent but not like the dog or the mobile hanging  over head.  And thus, as more adults are viewed, comparing their shapes and seeing similarities, this sameness permits creation – however nascent – of a thing or category via comparison and analogy.  Humans are like humans, but not like cats.

      As for love, it has always struck me that love involves a sharing of worldview, which might be described as resonating analogies.  So maybe that fits.

      • http://profiles.google.com/barry.kort Barry Kort

        Shared values is one basis for affective bonding, but another mechanism is propinquity.

    • http://profiles.google.com/barry.kort Barry Kort

      The brain is an analogy machine? I’m skeptical. This would mean that if there is a set of knowledge that has no analogy, then it will never be discovered. Ask yourself if this is true. You will need to prove this.  Or you can find counter examples to show that it is false.

      There is a particular class of mathematical knowledge, associated with apprehending the architecture and dynamics of recursive systems — systems governed by one or more recursion laws.

      These are notorious challenging systems to model with analogies.  Professor Hofstadter devoted his Pulitzer-Prize-winning book, Gödel, Escher, Bach, to this challenge.  In the end, his book was a narrative fugue, rather than a succinct free-standing analogy.

      I’ve struggled to find functional analogies for recursive systems, and while they exist, they seem exasperatingly ineffective at communicating the basic idea.

      See, for example, this effort to employ the analogies of Cancer and Whack-a-Mole:

      Recursion, Brain Cancer, and Whack-a-Mole

      • ToyYoda

        Thanks, I will look up the reference.  But, I can’t agree with Hofstadter having studied mathematics myself.  I did enjoy his regurgitation of the history of complex numbers, but he does not talk about the “spark”. Yes, complex numbers can be used like normal numbers algebraically, but what was the leap that last leap Bombelli took?  What about Newton’s discovery of the generalized binomial theorem?  What analogy was used there?

        The way I see analogies, is that they are tremendously useful.  They can allow for the tranmission of the “gist” of an idea to another.  They can lead you to a big picture thinking faster than other means.  And they can guide one to integrate new ideas into old frameworks (complex numbers).  

        It’s an expeditious framework for understanding new knowledge, but I don’t think it’s the most fundamental.  I just think it’s another tool in your mind’s toolbox.

        • http://profiles.google.com/barry.kort Barry Kort

          By “spark” I presume you mean making a new connection, out of thin air, in search of an appropriate analogy.

          These “sparks” can come from any number of sources. If I go for a walk in the woods, things I observe in nature might suggest an analogy. Sometimes they come to me in my dream sleep. Sometimes the key idea comes from something I am reading or hearing on WBUR or NPR. Sometimes they come from my recollection of a piece of literature or a long forgotten item of popular culture. Sometimes another person will suggest to me a usefull analogy to look into.

    • http://profiles.google.com/barry.kort Barry Kort

      How does a baby learn when he/she has no prior analogies?

      Most children learn from stories — nursery rhymes, fables, fairy tales, and parables. These stories are templates for abstract reasoning, where a child can imagine himself or herself in the role of one of the storybook characters.

      Even more interesting, how do we understand love? Faith? Happiness?

      Again, these concepts are typically conveyed through stories. The Princess Bride is exemplary in that regard. So is The NeverEnding Story (and many others).

      • ToyYoda

        Ah yes, stories, but what I say by nothing I mean nothing. A baby cannot not learn from stories unless he has already acquired sufficient enough language then the mechanism of analogy can take place.

        • http://profiles.google.com/barry.kort Barry Kort

          Before we have language, we (presumably) have vision and hearing.  Our dreams are perhaps the residue of visual thinking for most people once they have learned language. I note that I cannot remember any of my dreams unless I can construct a narrative account of the dream while it is fresh in my mind.  And what I am remembering is not the original (visual) dream, but the narrative account of it.

          Most of my dreams — even vivid and lucid ones — often prove impossible to capture in a coherent narrative.

          Hearing, I’m pretty sure, becomes attuned to recognizable patterns of sounds that otherwise are not (yet) language.  That’s why I believe the most important sound patterns are highly repetitive (like ma-ma and da-da), because these repetitive sound patterns become associated with life-sustaining aspects of baby’s life.

          Is it a coincidence that “ma-ma” and “da-da” have the same meter and also rhyme?  I suspect it’s not a coincidence.  So here we have an instance of resemblance which is a feature of analogy.

          Babies learn nursery rhymes which are simple, repetitive sound patterns with recurring meter and recurring rhymes.  Eventually a young child learns that those sound patterns correspond to words in a language, and these words are labels for something meaningful in the life of a child.  Note, also, that we tend to remember nursery rhymes until we eventually decode their deeper (semantic) meaning, after which we let go of the audionic “earworm” and remember the deeper message as a more succinct story.

          Think about it.  How many nursery rhymes that you learned as a baby (before you had language) do you still have rattling around in your audionic memory?  Try looking any of them up to discover what the experts say those cryptic nursery rhymes are really about.  Many of them encoded political taunts aimed at political figures of the day.

  • Jasoturner

    Should be a fascinating show.  I think Hofstadter and Sander are touching on a very important ingredient contributing to our ability to reason and think.  Even at a very unsophisticated level, the power of analogy to describe or characterize new things or experiences is obvious from everyday experience.  I can’t count the number of times I have explained things to my daughter by way of analogy with great success (at least, as far as I can tell.)

    I wonder sometimes if there might be a relatively and surprisingly simple mental “genotype” that delivers the phenomenally complex mental “phenotype”  that we see in humans.  After all, out base desires and ambitions are ultimately fairly few.  But their manifestation demonstrates infinite variety.  In this case, a small family of canonical examples might allow us to bootstrap a huge array of knowledge via analogy.

  • Gregg Smith

    If my brains were dynamite, I couldn’t blow my nose.

    I love analogies.

    • http://www.facebook.com/profile.php?id=1408098372 Mari McAvenia

      If my brain was a shotgun it would be better left unloaded.

  • Coastghost

    Ahhh, metacognition: how we think that we think! (or think that we think that we think)

    “Brain: the large piece of wax between the ears.”

    “Senses: the faculties of seeing, smelling, tasting, touching, and hearing that enable us to perceive the world without and equip us to hallucinate the world within.”

    • Dennis Lee

      Wow I just took about 18 sentences to try to say what you did in three. Am I being analogous?

  • burroak

    Curious, reading the Book title, I wonder if our brains are constantly, consistently comparing life-long data.

  • http://www.facebook.com/profile.php?id=1408098372 Mari McAvenia

    I get it. Great show.

  • ToyYoda

    Analogy works only from prior experience.  How does a baby use analogies to know the world if the baby has no prior experience?

    The guest uses the example that we know our mom, and from that we know  others have moms. But how do we know our mom in the first place?

    • http://profiles.google.com/barry.kort Barry Kort

      Got Milk? 

      “Ma-Ma” is the name of the Milk Supply.  Mammals have mammary glands, hence “Ma-Ma” means Milk. 

      And Milk also means Oxytocin Bonding.

      • ToyYoda

        Knowing mom is just an example, the question still remains. If analogy is the means of knowing our world as Hofstadter claims, then it means we have to be born with prior knowledge.

        I can go back even further using your example, how does the baby know *to label* “ma ma” with milk supply? Is it because everytime it says “ma ma” it gets milk? Well how does the baby know to make that sort of observation or even to correlate those two events? According to Hofstadter the baby had to have learn this style of correlation through means of analogy.

        Even more fundamental, how does the baby know that “ma ma” and getting milk are two separate events? How does the baby know to parse it’s environment so that these two things are separate events?

        • ToyYoda

          Sorry, what instead of “parse it’s environment” I mean “parse it’s stimuli”

        • http://profiles.google.com/barry.kort Barry Kort

          The question still remains. If analogy is the means of knowing our world as Hofstadter claims, then it means we have to be born with prior knowledge.

          Not quite. It means our brains must have the inherent capability of acquiring and reposing mental models that realistically capture, track with, and predict the salient and relevant characteristics of the world in which we find ourselves embedded. That inherent faculty is what make our species, Homo Sapiens, both sapient and sentient. 

          Initially, we acquire simple associations — recognizable stimuli (audio-visual patterns) that are variously appetitive or aversive (“good” or “bad”).  If Mama routinely utters the audible phonemes “ma-ma” when she picks up the baby for nursing, the baby will associate the visual image of the mother, the beating of her heart, and the utterance of “ma-ma” as sensory cues systematically connected to the joy of being nurse.

          A baby’s brain does this by means of statistical signal detection — picking out a recurring pattern from the jumble of background noise.  Our brains are evidently wired to detect these patterns and sort them into two distinguished classes, conventionally named Friend or Foe, respectively.  The evidence is that the Amygdala and Hippocampus are the lobes that are primarily responsible for this function.

          The faculty of Associative Memory is clearly a native faculty of the human brain. We learn to recognize the familiar, bonding to tribe or family by means of propinquity. Conversely we learn to become wary of the unfamiliar. And, as Freud noted, we become intrigued by uncanny stimuli which are <a title="I'm strangely attracted to you. | Ella Fitzgerald and Nat King Cole" href="http://www.youtube.com/watch?v=lEavX-Ciomg"strangely familiar and unfamiliar at the same time.

          • ToyYoda

            Fantastic display of knowledge formulated in a well digested manner. Thanks for this. But, I was hoping to set you up to say this. (And by the way, I would never be able to say this with the same alacrity.)

            If Hofstadter is to broaden the definition of analogy to encompass these rudimentary thoughts then okay, I guess Hofstadter is right, but he isn’t saying anything new. There is more information that is presente to us than we have synapses to save it all. So of course, there will be a mechanism of compression and reuse.

  • ccbard

    How is this book any different from George Lakoff’s book Metaphor and Thought from 1992?

    • ccbard

      It sounds like the premise is the same.

  • J__o__h__n

    Sometimes the brain sees patterns that cause it to draw false conclusions like intelligent design or confusing causation with correlation.

    • http://profiles.google.com/barry.kort Barry Kort

      There are some more compelling diagnoses to explain why individuals like Edmund Storms or Michael McKubre believe in falsifiable theories like Cold Fusion.

  • J__o__h__n

    Unless one wears toesocks, a mitten would be better for the analogy.

    • 1Brett1

      What, you mean you don’t wear toe socks?

  • http://www.facebook.com/simsobrien Chris O’Brien

    If we’re going to dive into our past and see how deeply analogies (and metaphors) go in our culture, you can look at the parables ascribed to Jesus in the Gospels. His descriptions of man’s relationship with God and with each other is all told through analogies and parables. You could say that, as a human construct, language is woefully inadequate to describe these metaphysical spaces between us and the infinite.

    • http://profiles.google.com/barry.kort Barry Kort

      I’ve long tried to find modern models for a number of insights that first turned up in theology, often with unfamiliar metaphors that one has to look up in scholarly sources. And indeed, I’ve found modern analogs of ancient wisdom. But it’s still next to impossible come up with modern analogies for these same ideas, mainly because the modern models are either recursive or mathematical.

  • Mark Felder

    Is there evidence of this in persons with brain damage?

  • justinjustout

     Music Sociologist  believe that when we hear new  music we weigh it against all the music we have heard thus far and we judge it based on that for its inadequacies.

  • PithHelmut

    Fascinating topic Tom. I just had something happen that illustrated the inhibiting internal guide we unconsciously live by; looking for a pen I used last night, I searched under books, under papers, on the floor, couldn’t find it at all but it couldn’t have gone far. Later after having given up and while doing something else, there it was, on the coffee table all on its own. This highlights what we do all the time – we expect things to be difficult so we don’t look at what is in full view.  We virtually live with our endowments shrouded, little do we realize. Our main preoccupation seems to be with living within an economy and in so doing, we simply abrogate our incredible magnificence. We reduce ourselves to just a notch above ants, willingly, with not a whimper of objection. 

    • http://www.facebook.com/profile.php?id=1408098372 Mari McAvenia

      Not ALL of us are willingly ignorant. Some have an intense drive to delve deeper into the mechanics of life. There are social sacrifices which come with the practice of thoughtful investigation. If one needs to think their own thoughts & express them, too, they must be prepared to take a lot of flack. “Blacksheep” are still infinitely more interesting than white ones to some of us, though.

      • PithHelmut

        So true. And change occurs when an organism (or a way of thinking) reaches “critical mass”.  I’d love to know what that ratio is.  Imagine – an epoch when black sheep become mainstream…

        • http://www.facebook.com/profile.php?id=1408098372 Mari McAvenia

          Then the white ones will be the counterpoints of the status quo. I like to “see” this idea rather than “think” it. I see a photographic negative. An inverted perception.

  • http://twitter.com/amsolom Andrea Solomon

    You can really notice this in children, i.e. when my daughter was very young 2-3 yrs old I was amazed that she could name animals from abstract representations (line drawings or silhouettes) if she was familiar with them from photos, or more realistic drawings of them, or seen them in real life.  Now at age 8, if she is trying to understand a new concept or situation, she often asks, “Is it like this…”

    • 1Brett1

      I lived in a commune in the 70s. I was a young single guy, but some of the people had small children. One girl (age 3 at the time) had never seen ice cream in a cone. However, she had seen lots of microphones (her dad was a sound engineer). She saw a cartoonish illustration of an ice cream cone in a children’s book and said, “look, a microphone!”  

  • Daniel Grant

    Wow!  What a treat to hear from Hofstadter!  I’ve read 2 books… finished “I am a strange loop” just recently.  I totally like the way this guy thinks:  kind, zany, insightful.  Though, I may not agree with everything that’s said ;).

  • http://www.facebook.com/profile.php?id=595452520 Cynthia Rose Osorio Florez

    I’m very interested in how this may present itself in the effects of alzheimer’s?? When my father was dealing with his disease before he passed, and now myself as I am getting older, I notice it is often some similar word or item that pops up in the brain when we are desperately seeking the ‘right’ word. But due to the seeming effects of the disease, not coming up with the right word.


    Well DUH. I thought everybody thought this way!

    But I guess that’s why I always seem to be looking at things differently than everyone else…

  • http://profiles.google.com/barry.kort Barry Kort

    My training in the STEM (Science, Technology, Engineering, and Mathematics) disciplines is in Systems Theory, Systems Thinking, and Model-Based Reasoning.

    As a scientist, I think with a models, ranging from explicit mathematical models to analogies, metaphors, parables, and even long form narratives.

    Umberto Eco said, “Where of we cannot express a theory, we must narrate a story instead.”

    Unlike Douglas Hofstadter, I’m fairly weak at the art of storymaking and storytelling, so I’ll typically have a technical theory (typically as an analogy) before I have any other model or narrative account.

    I did, once upon a time, undertake to write a piece in the style of Professor Hofstadter:

    The Turtling Test

  • ThirdWayForward

    It’s an interesting topic for discussion.

    What one needs are internal representations that are relational rather than based on atomistic features. Gestalt psychology posited mechanisms for organizing perceptual scenes in terms of relations between elements (a triangle is a triangle whether it is large or small, dark or light, colored or not, etc. etc etc). One also has the abstract concept of a triangle that can be transplanted onto other, non-geometrical situations, such as a love triangle or a political triangulation or a holy trinity.

    We think a theory of flexible relations also needs representations that are imbued with functions, purposes on a very deep level. All of the information that is represented in our brains includes an aspect that involves what that information is relevant for, in terms of our internal goals. This was a place where symbolic AI miserably failed (see Dreyfus, What Computers Can’t Do).

    Much of our learning of language involves analogical uses of words rather than literal interpretations. We are constantly adding new meanings to existing words based on analogical extensions.

    However, it’s not clear that the conventional computational models for neural networks can work for analogies, e.g. connectionist networks or simple feature comparators. They just don’t handle patterns in a relational way, such that a pattern can then be applied in some other domain.

    We need to rethink neural network theory from the ground up and the networks need to be embedded in adaptive, purposive, goal-seeking frameworks. When you look carefully at connectionist theory, it is really not well grounded in neurophysiology — there is a huge amount of hand-waving to connect the two, and it is very unclear that one can build a brain that does all that we (and animals) do in the flexible, combinatorial and analogical ways that we do it.

  • DeJay79

    Douglas Hofstadter maybe right or wrong I don’t know but what I do know is that if he wants to talk in interviews he needs to stop sucking breath in through his nose every other word! Damn that is annoying.

  • Pete Darnell

    I think these guys are on to something here. Look at how our entire legal system is based on precedent. Each new case looks for a previous similar case to decide on a ruling.  From politics: “As research has shown, the initial use of an analogy effectively sets the structure of the frame, and thus determines the sorts of inferences that people will draw.”  - http://scienceblogs.com/mixingmemory/2006/11/08/political-analogies/
    These guys are like Newton, everyone before him had seen apples fall from trees, but nobody put it together to note that it represented a deep underlying principle.

    • http://profiles.google.com/barry.kort Barry Kort

      Look at how our entire legal system is based on precedent.

      That is both a strength and a weakness of the Rule of Law.  On the one hand, precedent gives us a model for what to expect the next time a similar issue comes around. On the other hand, any mistakes are similarly replicated time and again.

  • Bob_in_Princeton

    This work by Hofstadter and Sander is great.  I’ve long viewed the brain as a pattern-recognition machine.  It starts with instinct – we (and ants) respond to an event or pattern, goes on to help us form the habits that allow us to get through the day without having to invent every action we have to take. 
    But this hurts when we don’t recognize that the pattern is somehow different: like an armadillo that avoids danger by jumping into the air, which is ill-served when the danger is an approaching automobile.  Or managers who view today’s problem as like last year’s and try to apply the same solution to what is really a different problem.

    And it’s necessary to learning and understanding, even when that learning comes because the known pattern doesn’t fit and we need to develop new ones, as Hofstadter mentioned with special relativity (an example I was going to give) or imaginary numbers.

    I’m looking forward to reading their book.

  • Josiah Vanvliet

    What are analogies an analogy for? What is the neural structure that you are describing?  

    • http://profiles.google.com/barry.kort Barry Kort

      Analogies are examples of models.  Other kinds of models are children’s toys, games, dramatic plays (with storybook characters), metaphors, and parables.

  • http://www.facebook.com/profile.php?id=595452520 Cynthia Rose Osorio Florez

    I’m also very interested in the guests thoughts about obvious [and not so obvious] patterns throughout our world. Are the branches of a tree ‘like’ the veins in our body or is what we are noticing more reflective of a fundimental structure in nature. So we are not using ‘analogy’ but simply recognizing the inevitable patterns repeating? We are, after all, a part of and not separate from our world, so what we create would therefore have to utilize those inevitable and fundimental patterns.

    • http://www.facebook.com/profile.php?id=1408098372 Mari McAvenia

      When I am admiring the naturally erratic patterns of tree branches in the wind I’m not thinking. When I take up a handful of jewelry findings and assorted stones, laying them out in a comfortable pattern- a satisfying array- I’m not thinking that the creation is an analogy for the tree branches. But, sometimes it is.

    • PithHelmut

      The fundamental patterns we observe are subject to interpretation and with our limited understanding, we could be chaining ourselves to them by not seeing the deeper and most essential concepts.  As we evolve, that picture will inevitably become greater.  Perhaps, if we could understand the patterns with greater clarity, we may learn to reach our illustrious potential. 

    • http://profiles.google.com/barry.kort Barry Kort

      The general structure of a tree is one of the most fundamental and recurring models in our culture.  So many things have hierarchical structures that resemble trees that the analogy or metaphor of a tree is found in many branches of discourse, and ubiquitously in technology systems.

  • http://twitter.com/aluhur Arthur

    what is the analogy behind the idea for the brain as the analogy machine?

    • http://profiles.google.com/barry.kort Barry Kort

      Model-Based Reasoning.  Analogies are exemplary models.

  • OliviaCruse1

    This seems a purely semantic argument, specifically an expansion of the term “analogy.” Are the authors simply saying that all knowledge is built on past knowledge?  If not, then what is distinctive about their view of “analogy”? 

    • http://profiles.google.com/barry.kort Barry Kort

      Try to find a counter-example of some new knowledge for which their is no prior metaphor or example.

  • ToyYoda

    I’d like to know if our guests think if computers capable of analogies?

    • http://profiles.google.com/barry.kort Barry Kort

      I have no doubt they would say, “Yes. Absolutely.”

  • http://www.facebook.com/sheron.burgess.10 Sheron Fraser Burgess

    If the brain is an analogy machine, what is intelligence?

    • http://www.facebook.com/profile.php?id=1408098372 Mari McAvenia

      The “chip”.

    • http://profiles.google.com/barry.kort Barry Kort

      Intelligence is the ability to think and solve problems.

      I often use analogies as a powerful tool for thought.

  • Dennis Lee

    I see it this way:  The idea that brain operates in an analogous fashion makes sense.  But what about the creation of values and judgement? 
    Our brains use the 5 senses as witnesses to the world we exist in.  The creation of the world and the story that each individual believes is his or her “reality” is an internal product of brain function.  The information that comes in through our senses; taste, sight, sound, touch and smell is sorted in a way to confirm and validate the ongoing story that is being crafted.  If there are 6 or 7 billion humans on the face of the earth, there are 6 or 7 billion stories running. The brain establishes “values” in this way: “this is good-this is bad”  “this is to be feared-this provides pleasure”  and so on.  As we grow, the human assembles his or her collection of values in an ongoing process throughout their lives and that becomes the story or the cosmology or the “It”.  When my wife says you don’t get “IT”, I think “Well, we get along pretty well but she has a different IT than I do and these differences in collected values are normal, essential to us all and the basis for judgement, both good and bad.  Each brain is put together differently but I’d better not fart under the sheets anymore.”

    • http://profiles.google.com/barry.kort Barry Kort

      What about the creation of values and judgment?

      See Insightful Learning.

  • Steve Levy

    The Myth of Metaphor, written in 1970 by Colin Murray Turbane, explored many of these concepts.  I had the good fortune to attend his philosophy course in 1973.  

  • Hugh Gerechter

    - It’s a shame the on-air discussion is caught up in the high-level analogies. Isn’t the starting point the inherent capacity for perception of shapes, motion, smells and sounds? 

    and – 

    Is ‘mapping’ a special case of analogy? I’m thinking about “Art of Tracking” by Liebenberg

  • Chip Hedler

    Just tuned in–I’d like to piggyback on ccbard’s question, “How is this book any different from George Lakoff’s book Metaphor and Thought from 1992?” I read Lakoff and Mark Johnson’s work, Metaphors We Live By (1980) about 30 years ago and would be interested in how you relate Lakoff’s work (which is all looking in from the outside) and your own thinking. And I just heard someone asking whether we in effect create our own reality based on our own experience, and a response that suggests that we need to find a degree of consensus in order to communicate and work together. That discussion echos the hairsplitting epistemological debates about consciousness found in many Mahayana Buddhist texts of two centuries ago.

  • ThirdWayForward

    The “poverty of the stimulus” argument for a genetic basis for language acquisition has a great many holes in it. To take it seriously you have to ignore all sorts of factors and shared attentional spaces between parents and infants.

    Analogical processing is likely to be a basic feature of neural networks.

    We have a very limited understanding of dog cognition. There was a NOVA program not long ago about the evolution of dogs and there was a border collie on that program that could recognize 300 different objects. Parrots use symbols, and can make up new ones. Let’s not make pronouncements about what animals can or cannot do from a state of ignorance.

    You can bet that if humans do it, at least some animals will also do it, albeit maybe not as well or not as often.

  • ToyYoda

    Hofstadter says that in order to make analogies, one has to get at the crux of the matter, and also to have a critical mind (I believe he said, “eliminate” possiblities.)

    I would argue that these two things which Hofstadter states “is needed to make analogies”, is in essence more fundamental than analogous thinking and more important.

    • http://profiles.google.com/barry.kort Barry Kort

      I agree. I routinely struggle to find analogies, not so much to aid my own thinking when I already have a functional scientific or mathematical model, but to come up with a way to explain the model to a lay audience that doesn’t apprehend the math or science.

      • ToyYoda

        But this is what I am getting at with analogies. They are a powerful that seems most useful in conveying complex ideas to others.

        Outside of every day things, I find that they show limitations fast. I have read countless popular books on astronomy and quantum physics and have come across several passages whereby the scientist say that there is no good analogies.

        In my own field of study, mathematics, I don’t come to know something by way of analogies. It’s easy to give a student a problem where analogies would be difficult to use. Countless math olympiad problems in geometry would suffice where clever inventions and key observations are often what’s needed. I have to do this every day, which is why I am taken aback by Hofstadter’s claim that analogies is the major way in which we learn about the world.

        • http://profiles.google.com/barry.kort Barry Kort

          “Whereof we cannot express a theory, we must narrate a story instead.” ~Umberto Eco

          Only a tiny fraction of the population ever learn or use advanced methods in mathematics. The vast majority of people, who have negligible access to the art and practice of mathematical modeling, are obliged to rely on metaphors, analogies, parables, games, dramas, and narratives instead.

      • ToyYoda

        I too have trouble giving analogies to others, but I don’t have to worry about grants or maintaining a good public relations as scientist have to. So, I do it far less often and with very little consequences. As you can see as I grope for what I know intuitively to be right regarding the use of analogies.

        Anyways, I’m not shy to use analogies as a tool to understand and to explore, but I often prefer to know a subject on its own merits and am aware that analogies can only go so far, and in the worse case lead you astray.

        Actually, I more often use analogies as a tool for verification. So if you were to explain to me your ideas, I may interrupt you to make an analogy or to paraphrase what you have said. And then I would be curious to know what you thought of my analogy. If you are favorable, then I know that I am heading towards the gist of your ideas, while “internally”, I am am trying to understand your ideas on their own in terms.

  • Kamesh Aiyer

    In the seventies, Newell at CMU did some work on discovering analogies through pattern matching. The conclusion was that analogy had to be built into the architecture of cognition and that there had to be a base set of “concepts” built (by our DNA, presumably).

    The implication of that is that whatever analogies our brain uses, we won’t be able to discover them and that Chomsky is partially right — there is something built into the hardware (and wrong because this is not purely human).

    The question then is “How come people report using analogies to make discoveries”.  The keyword there is “report”.  “Reporting on the inner state of the architecture of cognition” is a problem to be solved by analogy! So what we get when Einstein reports sitting on a lightbeam is not the analogy his brain used but the analogy that the brain made in order to explain the analogy it actually used.

    I think we call this rampant re-ification.

    • 1Brett1

      You touch on something interesting there. We’ve all had experiences/feelings we’ve never experienced/felt before and yet have drawn reasonable conclusions from them but not based on analogies, associations, etc. I remember when I was very, very young having certain feelings about my perceptions of space and time (without any prior experience with/knowledge of concepts regarding this phenomenon). I later remembered those thoughts in reading about certain scientists/their discoveries/their theories. I remember wondering what compelled them/gave them the impetus to explore certain ideas without any prior model/analogy/association, etc., to work from. Did they start with some undefined feeling/experience similar to mine then were driven to get at potential possibilities?

      • http://profiles.google.com/barry.kort Barry Kort

        My most commonplace emotion, when encountering something unfamiliar is bewilderment, followed by perplexity

        If, in attempting to resolve my perplexity, I begin to form an inchoate (and almost surely incorrect) theory, I next tend to experience confusion, because my initial mental model doesn’t seem to jibe with my observations. 

        If I labor in vain to apprehend a coherent and reliable theory or working hypothesis or mental model, I am liable to arrive at a state of frustration, exasperation, or despair.

        At that point, I might conclude the puzzle is too intractable for my feeble brain, and set it aside for the time being until I can do some more research or literature review. And if that fails, and the problem is of sufficient importance, I might also seek a subject-matter expert, if one can be found.

        At that point, my feelings of hopelessness and despair might turn to hope that providence will someday provide the long sought insight and solution to the bewildering and perplexing puzzle.  If and when that happens, I might become surprised, astonished, gratified, or elated.

        • 1Brett1

          In reading your reply, which I admit I didn’t finish because you missed the sort of feeling I was getting at, it was obvious you read my comment with a ‘commonplace’ experience in mind. 

          I remember having a unique thought as a child, never experienced before, and feeling perfectly comfortable and with a sense of wonder. E.g., ever feel as though time stood still? Ever remember as a child observing/sensing the relativity of time for the first time for each person in each moment? Ever remember seeing another for the first time as anxious and feeling he/she has no time because he/she is on his/her way to work or to an uncomfortable task, yet you had all of the time in the world….and a light bulb went off in your head where no analogy or model existed? ever remember a feeling of everything is in it’s place and prefect? That time had elasticity? If so did those feelings cause you bewilderment? Confusion? That you were beset with an intractable puzzle? Frustration? exasperation? Despair? They didn’t for me.

          In your reply, you seem to be spending your reply explaining the gamut of uncomfortable feelings that eventually get resolved rather than attempting to see a “oh, I get what he is trying to convey,” which is disappointing. Kamesh Aiyer seemed to be “arguing” a point for the sake of it by building straw men out of others comments so he could knock them down even “arguing” for its own sake. You were taking a different tact. But, in the end, you both are advocating for a kind of “no, that is wrong, my ideas are right” kind of approach to discussion. I doubt that kind of talk of how the brain works is very interesting when reduced to different camps or “I have found the way.” 

          Besides, you seem to be selling something. I didn’t listen to the show; I thought it probably would have been too much of a “we now can explain how people think” type deal to be interesting and it was trying to sell something, i.e., books. What you’re selling isn’t as ostensibly whorish but perhaps it pertains to having people mosey over to your blog?   

          • http://profiles.google.com/barry.kort Barry Kort

            First, please understand that I was responding to both Kamesh and your response to him.  

            DISQUS, alas, doesn’t provide a way to label the scope of a rejoinder.

            I don’t claim to know or understand how cognition, affect, and learning works for any given individual.  At best I have a model that seems to be able to cover the spectrum, more or less.

            The only specific example I can offer here is my own idiosyncratic experience, within the scope of that model.

            I have, indeed, experienced the elasticity of time, which I suspect is a feeling mediated by tachykinins (or some related neuropeptide) which evidently has the ability to modulate the firing rate of neurons.

            Dolly Parton performs a nostalgic song entitled, “Do I Ever Cross Your Mind,” which establishes a very regular metronomic beat, and then unexpectedly inserts a stretched beat on a particular lyric (“How … I wish we could go back in time”) that gives the listener the sensation of time stopping for an instant, then speeding for two beats.  It’s an utterly uncanny listening experience if you don’t know it’s coming. (Listen, also to the studio album version, where the time-warp effect is optimized.)

            What I’m selling (actually giving away free of charge) are my own insights, which readers are welcome to cherish or discard, as they see fit.

          • 1Brett1

            I did listen once…I’d have to listen a couple of times to be sure but it sounds as though the beat isn’t stretched at all (no irregular, syncopated measure is mixed in with regular measures). Perhaps you are confusing the length of the beat in a measure with the length of a note or couple of notes within a measure (two very different concepts). What also adds to the dramatic sense of “hesitation” is the instrumentation rests, or stops briefly, behind her voice.

  • Kamesh Aiyer

    The work that Newell did in the seventies culminated in the 1980s with an “architecture for cognition” called SOAR.  The technique used by SOAR for analogies was to initiate a search for a solution pattern (an existing goal or problem-space or operation) when problem-solving reached an impasse.

    Note again that this happens in the context of problem-solving. The brain does nothing when there is no “problem” to be “solved”.  That includes the discovery of the foot:hand::sock:glove analogy (i.e., if gloves or socks were actually created by analogy).

    As presented in the interview the foot/hand analogy is not an analogy the brain used, but one presented as a problem to be solved.  It is almost certainly NOT the analogy used inside your brain!

    • Wm_James_from_Missouri
    • 1Brett1

      I didn’t listen to the show, so you lost me at, “…discovery of the foot, hand; sock, glove analogy.” I’m not sure of the context you have used/why you have used the word “discover”?  We didn’t discover analogies but invented them. As far as making comparisons, it’s difficult to say why humans tend to make comparisons or associate one thing with another as a mechanism of brain function/activity. 

      As far as ways we use our brains to solve problems. It seems we either use our brains to solve problems or we are using them for perseveration, depending on one’s definition of “solving problems.” When we express emotions (either through verbal or nonverbal communication), in a sense we are solving problems (the problem of needing/wanting to communicate). If we daydream, we are usually creating some fantasy world. In that world, as we daydream we are creating various defined elements to the fantasy world. This is also a form of problem solving. Even in sleeping dreams we are problem solving, we just do not employ the same editing processes as in our waking state.

      Now, if we are worrying, where we move our thoughts around in a circular tape-looping manner, we aren’t problem solving but perseverating.

      Perhaps when we are meditating (if we are doing it properly) we are doing neither problem solving nor are engaging in perseveration, although a mantra is a kind of purposeful perseveration that is soothing. 

    • http://profiles.google.com/barry.kort Barry Kort

      The brain does nothing when there is no “problem” to be “solved”.

      Boredom is the mother of adventure.

      If I don’t have something to engage my fascination, I’ll go looking for something, such as recreational puzzle-solving.

  • 1Brett1

    Of the dogs I’ve owned in my life, they seem good, attentive and vigilant observers. At times, however, they’ve had difficulty interpreting their observations. If they run across something they’ve never seen before, they do not generalize information stored from their past or new information to determine what it is they are looking at. They can’t generalize new information or make analogies from similar, old information. This is what separates, I believe, the human brain from the animal brain. 

    • Wm_James_from_Missouri

      Most dogs have a problem with the concept of ” I am pointing to … ” .

      • 1Brett1

        Heh…the concept of trying to see what we see probably doesn’t occur to them. Also, there is that, “ooh, is there food in the hand with that pointing finger?” 

      • 1Brett1

        Another thing funny about some dogs…when they play with us, say, with a ball, their behavior suggests that they think we are as excited about the prospect of possessing/chasing the ball as they are…It’s cute/endearing. 

        • Wm_James_from_Missouri

          You are so right ! All of my dogs have loved going after that ball. Most of them loved to play hard to get (it back). The very fact that we are paying attention to them is enough for them. Kind of like some women I have dated. ( I just had to say it.)

  • Kamesh Aiyer

    Brett, you’ve just cited a “flawed animal cognition field study.”  There was a recent article (in the Wall Street Journal, of all places) on other famous flawed studies. (see http://online.wsj.com/article/SB10001424127887323869604578370574285382756.html).

    The point is that we humans have no idea of how to assess the cognitive abilities of animals and the more we do it on the basis of some expertise we possess, the less valid it becomes.

    • 1Brett1

      I didn’t cite anything; I was making my own (unscientific) observation and opinion. There is no way we can know, for sure, how animals process information. 

      I’ll have to look at the link (thanks). I find it interesting that a dog, say, does seem to respond to stimulus he/she has never seen before. The response is very different from, say, a human with a fully formed brain. The dog’s response is more like a very, very small child (under 1 year old, say), as if magic is happening or something very ordinary/pedestrian is met with fear, apprehension, and so on.

    • 1Brett1

      Sorry, your link said, “page unavailable.”

    • 1Brett1

      I did just find an article in the WSJ that may have been the one you were linking…My point wasn’t to suggest human exeptionalism, just that there appears to be a difference between the way humans and animals use their brains. 

    • http://profiles.google.com/barry.kort Barry Kort

      The point is that we humans have no idea of how to assess the cognitive abilities of animals and the more we do it on the basis of some expertise we possess, the less valid it becomes.

      How do you falsify the assessments of Temple Grandin or Irene Pepperberg?

  • dcmcguire

    If analogies create generalized concepts, which produce unconscious behavior, how do long hours of violent video gaming integrate into human thinking, associations and behavior?

    • 1Brett1

      I didn’t hear the show…but I don’t see this idea as an attempt at teaching/learning conformity, in terms of making one teaching technique fit all. 

      Can you cite examples where humans do not use generalizations, comparisons, analogies, or associations etc., to interpret new information/remember old information? Even with outmoded teaching/learning approaches such as memorizing information in rote fashion, the brain is making associations. 

      • dcmcguire

         I’m curious about the role of repetition, video games and otherwise on cognitive and genetic functioning.  How are analogies coded by the brain and DNA.   Do analogies link to form and strengthen neurological pathways and potential epigenetic influences? Does repeated thinking and behavior seek conforming analogies, so that an exponential influence is exerted on neurological and epigenetic signalling (methylation or acyetalation) ? Any thoughts?

        • 1Brett1

          I play a couple of musical instruments and have done so since I was six (58 now). I also teach music. I came up with three functions of teaching/learning (sometime during the process of listening to a student play a new piece of music one early Saturday morning). Imitation, repetition and analysis. It seems if a student doesn’t use each function of these three, he/she doesn’t learn very effectively. 

          In my lessons I use all three in every lesson. In my introduction to a given lesson I demonstrate what I want the student to learn and get the student to imitate my demonstration. I then engage the student in a little analysis through going over a little bit of music theory and asking him/ her questions, and I have him/her ask me questions. I ask him/her to repeat the demonstrated lesson over and over in his/her home practice; I also ask him her to come up with three questions at home about the lesson to be asked when he/she returns the following week. 

    • Kamesh Aiyer

      More like, the establishment of an analogy in the brain IS the epigenetic event.

      • http://www.wiremybrain.com/ donna-christine mcguire

         Well put.

    • http://profiles.google.com/barry.kort Barry Kort

      One of the best sources of an answer to your question would be to view the “Let’s Play” series by Sven Groot on YouTube.

      He records himself playing a wide selection of video games (including both non-violent puzzle-adventure games and violent games like Mass Effect).

      Mr. Groot narrates his game play in a delightfully engaging manner, carefully explaining his thinking as he goes along.

      • http://www.wiremybrain.com/ donna-christine mcguire

         I will indeed take a look.  The concept sounds interesting. Really appreciate your suggestion. Thanks so much!

  • 1Brett1

    It seems much of the commentary on this forum has been in the realm of something binary or absolutism. I haven’t heard the show, so I can’t comment on that. I hope it didn’t present its ideas as our brains have to either work this way or that way, and the theory talked about is it works this way and not that way. I believe that different parts of the brain function differently, and that what seems to make one part function a certain way may explain in part how it works but it doesn’t explain absolutely how it works. Maybe I’m too gestalt… 

  • http://profiles.google.com/barry.kort Barry Kort

    If you look at how human languages work, they’re not very literal but, instead, full of idiomatic expressions.

    Gidoudda town.

  • Kamesh Aiyer

    The core problem with Hofstadter & Silva’s proposal is that in the form they discussed it in the show it is not falsifiable.  The analogies they point to and the analogies that their opponents point to are all reported (by the mouth, so to speak) by the thinker and not observed by a third-party.

    If you agree with Popper, this proposal then is no longer scientific.  The thing that would make such a proposal scientific is a defined “architecture of cognition” that can be a) shown to use analogy effectively and b) is cognitively plausible.  But, other than the SOAR work that I cited (and it is over 30 years old now) that isn’t the way the field has gone.

    • http://profiles.google.com/barry.kort Barry Kort

      I suppose it can be shown that some individuals lack the ability to reason in general, and by analogy in particular (presumably because of some brain damage), but almost anyone who is not brain damaged can be taught to reason by analogy.  So what would it mean to falsify it?

      • Kamesh Aiyer

        The hypothesis is “The brain solves new problems by making analogies to old problems that it knew how to solve”.  To falsify it would be to show that the brain does NOT use analogy to solve new problems.
        This is not the same as “reasoning”.  When I give you my reasons for doing something, I am reporting, not thinking.  Thinking is what goes on in the brain, reasoning is the thinkers explanation.  
        Almost by definition, this cannot be what the brain is doing.
        If the argument is that the brain is doing that at a “higher”, “conceptual” level, then ANY low-level neural process can be described at some higher level as making an analogy.

        H&S did exactly that on the program when people brought up scientists who have made sui generis discoveries — they showed, for example, how Einstein imagined what he would see if he were sitting on a beam of light.  There is no way to prove H&S wrong.

        Having said that, I would like to know what analogy Newton used for the law of gravity (the inverse-square law is the only one that satisfies  Kepler’s discovery of elliptical orbits, but how do you get there by analogy?)

        • ToyYoda

          Bravo. Nicely said. I wish I could be so clear! I do have issues with Hofstadter’s example of Einstein. I think it’s quite a stretch to consider these Gedanken experiments analogies. The analogy is not the method of understanding or discovery, but rather the end result of that thought experiment, e.g., that constant acceleration in an elevator is similar to gravity on the surface of a planet.

          • http://profiles.google.com/barry.kort Barry Kort

            I suppose a Gedanken experiment is the equivalent of an author constructing a narrative story, drama, or play and imagining how it would play out. This is a bit like a simulation model, except the simulation is running in a human brain instead of on a computer.

            I’ve on occasion written whimsical parables or parodies as an alternative way of presenting the essence of a technical idea.

          • ToyYoda

            We can construe stories, dramas, plays and fables to be Gedanken experiments.  But there’s a reason why stories, dramas, and plays are not called analogies, because they aren’t!

            I guess I take issue with Hofstadter’s broad definition of an analogy that I wonder if he is making a cognitive mistake, or setting up his unsuspecting reader for a cognitive trap.  Perhaps there is a term for it….

            Let me explain the error by way of analogy (oh the irony!!).  I’ve run into people who believe the world is a struggle between the Haves and the Have Nots.  Other optimistic people believe that the world can be described as  love and hate.  And then there are gold bugger who believe the world is a string of currency debasing government conspiracy theories.  Since these are just view points, any situation can be coached (spun) into these views.  But they are just view, and we can adopt another equally valid viewpoint for understanding events.  Then there is the additional mistake of turning these viewpoints into something ontologically real.  So there is 3 things.  A view attachment.  Spinning everything to fit that view, and then believing it’s ontologically real and since it’s a viewpoint, it’s easy to spin all events in its terms which only strengthens their believe that the view is real.

            I wonder by saying everything that we understand is by way of analogy, that Hofstadter could be roped into such a thing.  He certainly demonstrated it with the exampe of complex numbers.

          • http://profiles.google.com/barry.kort Barry Kort

            We can construe stories, dramas, plays and fables to be Gedanken experiments.  But there’s a reason why stories, dramas, and plays are not called analogies, because they aren’t!

            Stories, dramas, plays, fables, and parables are all narrative varieties of Model-Based Reasoning.  Many of these story-based methods do incorporate metaphors and analogies within them.
            Probably the best known examples of parables with embedded metaphors and analogies are the ones commonly used in theology.

            So, if I survey Model-Based Reasoning, I can start with the narrative formats, then highlight the judicious use of metaphors and analogies, and finally move on to games,  simulations, scientific theories, and mathematical models.

            When I am working with scientific theories and mathematical models, I usually have to struggle to find good metaphors or analogies that translate highly technical models into something a lay person would recognize and appreciate.

            Consider, for example, a problem that Doug Hofstadter is noted for: finding a novel way to devise narratives, metaphors, and analogies to translate the arcane concept of Mathematical Recursion into something a lay audience can digest.

            Here’s another one to ponder: Find a good way to explain the mathematical concept of a function to a lay person, using only metaphors or analogies. Most lay people are quite familiar with the word ‘function’, but not with the technical abstraction that word refers to in mathematics.

            Or, if that’s too hard, try to find a metaphor or analogy to explain the concept of an abstraction!

            Hint: You might need to turn to a joke, anecdote, cartoon, or piece of abstract art.

        • http://profiles.google.com/barry.kort Barry Kort

          The hypothesis is “The brain solves new problems by making analogies to old problems that it knew how to solve”.

          There is a missing quantifier in the statement of the working hypothesis.  Let’s consider three variants of the working hypothesis by inserting one of three possible quantifiers:

          The hypothesis is “The brain solves new problems by {sometimes | always | never} making analogies to old problems that it [already] knew how to solve”.

          It’s clear (at least to me) that we often (or at least sometimes) turn to analogies. That also falsifies the “never” case.  

          The only one that’s worth falsifying is the “always” case and I can attest that I have wrestled with problems for which the solution turned out to be a novel technical theory or model, for which no viable analogy has yet surfaced. 

          It’s not that I cannot find some analogy, but I find that when I use some plausible analogy with a lay audience, their eyes glaze over (at best).  

          And, in the worst case, their eyes bleed, their head asplodes, and they quickly run me out of town on a rail, with screams of exasperation wafting from all quarters.

          • Kamesh Aiyer

            Barry, when you said,
              ‘It’s clear (at least to me) that we often (or at least sometimes) turn to analogies. That also falsifies the “never” case.’
            You are making exactly the error I am pointing out — you are confusing “we” with the “brain”.  
            Then, when you say,
              ‘I can attest that I have wrestled with problems for which the solution turned out to be a novel technical theory or model, for which no viable analogy has yet surfaced.’
            You repeat the error. What you can attest to, or report, is a construct that your brain created when asked “what is the analogy?”.  The brain can easily fail to report any analogy it may have used because the  low-level mechanisms used by the brain are not accessible for reasoning about.

            Consider whether you can possibly report the following:
              ‘I first tried to solve the problem with the left side of my brain and then realized that I needed some new ideas, so then the right side of my brain came up with three ideas, but the left side rejected two of them….’
             ‘I had saved my solution to the problem P as a collection of microRNAs stored in neuron numbers 3,321,785-3,231,993 and also in… Then when faced with problem N, I used the ATP cycle in neuron 3,355,787 to activate these neurons…’

            (It’s a rhetorical question,please don’t answer it).

            The thesis that every cognitive act can be DESCRIBED using analogies is completely UN-interesting.

            What you (or anybody else) can “attest to” does not prove anything because you cannot possibly know how your brain did something.

            But H&S repeatedly such testimony to bolster their claim that the brain thinks using analogy.

            Then when you try to pin them down with specific situations (“How does a baby learn a concept”), they point to long chains of analogy that make it impossible to answer such questions except in the most general way.

            The amazing thing is that H&S have managed to write a fat book based on a simple, well-known, and much-discussed, categorical error of the difference between the “symbolic” and the “physical” levels at which the brain functions.

          • http://profiles.google.com/barry.kort Barry Kort

            Kamesh, when I say,  “I can attest that I have wrestled with problems for which the solution turned out to be a novel technical theory or model, for which no viable analogy has yet surfaced,” I am talking about a theory that someone else created, and which subsequently came to my attention.  Most scientific theories that I become aware of were developed by someone else, and turn out to be the missing model for which I previously had nothing better than the Null Model up to that point.  If you like, I can give you some examples of full-fledged theories that I stumbled upon after someone else published it, and for which no viable analogy exists (meaning that the analogies I have tried leave people staring blankly at me).

            What you (or anybody else) can “attest to” does not prove anything because you cannot possibly know how your brain did something.

            I can attest, with complete confidence, that my brain utterly failed to come up with anything even close to a viable model, and then someone came along and simply gave me the solution that had utterly eluded me.

            I do a lot of work with recreational puzzles, and although I’m better than most people at discovering a solution, I can assure you there are puzzles which had me utterly stumped.

  • http://www.facebook.com/anthony.browness Anthony Browness

    Many of the most powerful and simplistic Programming concepts are analogous.

    An enhanced for-loop uses terms such as “FOR this AS that”.
    If-statements are built like “IF this is equal to that, or greater than this, and like this THEN do that”.

    Abstraction deals with parent classes that share traits with children classes.

    Do your guests think we built computer concepts to think like we do, or they just ended up that way because analogy is sometimes the best way to go?

  • http://www.facebook.com/arenbaum Peter Baum

    There are other theories about thought that may be of interest.  My own theory, “ThoughtForms,” is described in my recent book.  Although reasoning by analogy is discussed at the end of the volume, that strategy of thinking is just one of many that people use.  In other words, the basis of thought, according to ThoughtForms, is quite different from what Hofstadter and Sander suggest.

    ThoughtForms theory developed from artificial intelligence research, although the theory has much broader application.  The book is very specific, defining thought explicitly, using the theory to define ideas like numbers, and showing the implications of the theory for a wide range of subjects. The theory also has predictive power, for example, suggesting which ideas and beliefs are likely to be controversial and
    why.  ThoughtForms’ implications for learning and teaching also are described.  In addition it has applications relevant to creativity, such as explaining why music is universal.  The theory also can be used in the creative process itself, for instance, in the invention of new kinds of numbers.

    ThoughtForms is a widely available E-book and was recently published as a paperback (available through Amazon).  The title of the book is ThoughtForms-The Structure, Power, and Limitations of Thought (Volume 1-Introduction to the Theory).

  • Adrian_from_RI

    Tom, are your guests proposing a solution to
    the problem of the universals? If so, then they might be interested in reading
    the book “Introduction to Objectivist Epistemology” by Ayn Rand.  

    • http://profiles.google.com/barry.kort Barry Kort

      Adrian, does Ayn Rand have model of Learning Theory, revealing how a learner proceeds from a vague, inchoate, or incomplete understanding to construct an accurate, reliable, and complete knowledge base?

      How, for example, does she model the process of diagnosing gaps and misconceptions that inevitably find their way into one’s partial understanding of something during the middle phases of a learning journey?

      • Adrian_from_RI


        My field was electrical engineering and only since my retirement
        have I taken an interest in philosophy, in the world of ideas. Therefore, I am
        not knowledgeable enough to answer your question. But listening to Tom’s guests
        I was wondering whether their study of how our mind works was enlightened or
        not by an understanding of – what is known in philosophy – the problem of the
        universals or the problem of concept formation. 

        This problem was first exposed by Socrates (469-399BC)
        and the inconvenient questions he confronted the Athenians with. Plato
        (428-348BC) started looking for answers in the supernatural World of Forms, the
        world of Ideas. I call that world La La Land. Aristotle (384-322BC) disagreed
        with his teacher Plato. The One in the Many, the concept was not in La La Land
        but must be here in this world. I found it fascinating to learn about men’s
        quest for knowledge, our quest for understanding the world we live in; to learn
        of men’s 2,100 year slow progress from Aristotle to the founding of America in
        the Age of Reason; and from that pinnacle down to the present Post-modern era
        marked by skepticism and nihilism.  “The
        History of Philosophy” lecture series by Dr. Leonard Peikoff proved priceless
        and indispensable in learning about the world of ideas. You might want to down
        load these lectures from: 



        Maybe this Onpoint program “The Brain As An Analogy
        Machine” was of interest to you in your field of Affective Computing Research.
        But I think it might also be useful for you to know the answer to the problem
        of the universals. And that is why I recommended to Tom and now to you to take
        note of the book “Introduction to Objectivist Epistemology.” The first time I
        read that book many, many years ago I found it boring and rather
        commonsensical. But that was before I knew about the problem of the universals;
        a problem that is now causing the world to go to hell in a hand basket.



  • chanankub

    The gentlemen present a nice analogy
    of the human mind. Their discussion of the mind is analogous to Hofstadter’s
    description of looking for the elevator in a strange environment: they take all
    they think they know about the mind, toss out parts that cloud the picture,
    come up with something familiar and voila! That’s the mind!

  • The_Truth_Seeker

    Funny, I was just thinking about what happened to Douglas Hofstadter a few days ago, before I heard about this program… I have both his earlier books and wondered what he was up to. Was waiting for him to develop the first thinking computer. 

  • The_Truth_Seeker

    When I was 16 (many years ago) I became convinced that thought STARTED with the emotions and evolved from there (rather than originating from language and abstract logical understanding). This was at a time when most researchers thought it could be reduced to logical rules and language (I never believed this). Thought evolved from basic “feelings” and sensations and then the evolution of more complex emotions, memories and recognition of important patterns. I also build an optical “perceptron” that could reduce 2-shapes to a single number (regardless of orientation but not made public at the time). Hope to get back to some of this earlier work … someday.

  • The_Truth_Seeker

    With all this work on AI, how come we can’t explain self-consciousness yet and how and when that “happens”? Seems like that’s a pretty fundamental question and test of AI.

  • The_Truth_Seeker

    They are starting with pretty abstract concepts. This has been the problem with AI all along. It starts with much more primitive elements but they include some forms of pattern recognition (but maybe not analogies). Analogies have to come much later in the process, I think (just like language and logic).

  • Bibliodrone

    I can see why Jason Silva calls his video an “espresso shot” (nice…analogy). Waiter, I’ll have what he’s having!

  • Pingback: The Brain As An Analogy Machine | franzcalvo

Aug 29, 2014
Ukrainian forces guard a checkpoint in the town of Mariupol, eastern Ukraine, Thursday, Aug. 28, 2014. Ukraine's president Petro Poroshenko called an emergency meeting of the nation's security council and canceled a foreign trip Thursday, declaring that "Russian forces have entered Ukraine," as concerns grew about the opening of a new front in the conflict.  (AP)

War moves over Syria, Ukraine. Burger King moves to Canada. Nine-year-olds and Uzis. Our weekly news roundtable goes behind the headlines.

Aug 29, 2014
Beyoncé performs at the 2014 MTV Music Video Awards on Sunday, August 24, 2014 in Inglewood, California. (Getty)

Sex, power and Beyoncé’s feminism. The message to young women.

Aug 28, 2014
Some of the hundreds of earthquake damaged wine barrels cover and toppled a pair of forklifts at the Kieu Hoang Winery, Monday, Aug. 25, 2014, in Napa, Calif. A powerful earthquake that struck the heart of California's wine country caught many people sound asleep, sending dressers, mirrors and pictures crashing down around them and toppling wine bottles in vineyards around the region. (AP)

Drought in California, earthquake in Napa. We look at broken bottles and the health of the American wine industry.

Aug 28, 2014
Photos surround the casket of Michael Brown before the start of his funeral at Friendly Temple Missionary Baptist Church in St. Louis, Monday, Aug. 25, 2014.  (AP)

The message that will last out of Ferguson with New Yorker writer Jelani Cobb.

On Point Blog
On Point Blog
Drew Bledsoe Is Scoring Touchdowns (In The Vineyards)
Thursday, Aug 28, 2014

Football great — and vineyard owner — Drew Bledsoe talks wine, onions and the weird way they intersect sometimes in Walla Walla, Washington.

More »
Poutine Whoppers? Why Burger King Is Bailing Out For Canada
Tuesday, Aug 26, 2014

Why is Burger King buying a Canadian coffee and doughnut chain? (We’ll give you a hint: tax rates).

More »
1 Comment
Why Facebook And Twitter Had Different Priorities This Week
Friday, Aug 22, 2014

There’s no hidden agenda to the difference between most people’s Facebook and Twitter feeds this week. Just a hidden type of emotional content and case use. Digiday’s John McDermott explains.

More »
1 Comment