The Ashtray Argument

Mar 07 2011 Published by under From the Melodye Files

"I call Kuhn’s reply “The Ashtray Argument.” If someone says something you don’t like, you throw something at him. Preferably something large, heavy, and with sharp edges. Perhaps we were engaged in a debate on the nature of language, meaning and truth. But maybe we just wanted to kill each other."

--Errol Morris in the opinion pages of yesterday's NY Times, aptly capturing that spirit of wonder academics bring to the world... Or something.


20 responses so far

  • Avery Andrews says:

    I'm mildly relieved to see that Kuhn is revealed to be a lunatic, because I think the paradigm idea is the beginning of the path that leads to science-denial as forex discussed just today here here: Chomsky was supposed to a paradigm shift from the structuralists, but I was one of the first group of students taking Paul Kiparsky's version of the old 'bad guys' course at MIT, in which they were good guys with a bit less in their toolbox.

  • daedalus2u says:

    The paradigm shifts in science are exactly right. Read the quote by Barbara McClintock.

    That is what she came up against. The work-a-day scientists couldn't understand, couldn't even see what was obvious to her.

    If your brain doesn't have the neural network to instantiate an idea you can't even think about the idea. You need to modify your neural network until it can, and that is not always easy because of hysteresis.

    That causes an inability to understand new ideas because they are incommensurate with the old (what I think that means is that the new ideas don't "map" onto the same neural network the old ideas already map onto. You have to change the neural network to get the new idea to "fit").

    That is the essence of conservatism, not being able to see and understand new ideas. The GOP can't understand AGW because it is incommensurate with a 6,000 year old Earth and YEC.

    • Avery Andrews says:

      There's something to this, but it's also clear that not everybody has this kind of problem, you can grow new networks, as it were, by putting some effort into it. Of course many people are too busy with other stuff to put in enough effort, and others don't know that there's stuff they need to do.

      There's also an issue of whether it is in-principle incommensurability, or the difficulty of shifting to work with different complex concepts. I suspect the latter.

      • daedalus2u says:

        There is no distinction that can be made between "in principle" and "too difficult for some people".

        Essentially everyone does have this problem until they do grow new networks. This is why it takes time to understand something, and why education requires a path; starting with the basics and working to the more complex.

        If your neural networks can't instantiate an idea, you are as ignorant of that idea as it is possible to be. You don't understand it, you can't understand it, you can't even think about it or recognize that you are ignorant of it. The usual practice is then to dismiss the idea using the arrogance of ignorance.

        • Avery Andrews says:

          If you put in the time, you might understand some new idea. Whereas the paradigm is widely understood to be something you can't get out of. The fact that 'hard' has an extensional resemblance to 'impossible' makes it even more important to distinguish them.

          I wouldn't bet that it 's just a matter of 'growing new networks', although my time-scale for learning new math is consistent with something like that happening (stare at paragraph today, mind reels in baffled incomprehension, next day, a glimmer of light, perhaps, some time later, a bit of understanding). There are plenty of other reasons why somebody might reject a new idea.

          Well I myself might be an example of paradigm-shift failure, having never made the shift from the Aspects of the Theory of Syntax/Remarks on Nominalizations view of Generative Grammar, to the Conditions on Transformations/GB/MP one. Justifiable skepticism, or something less complementary? Posterity will have to judge that one.

          • Fred says:

            "Well I myself might be an example of paradigm-shift failure, having never made the shift from the Aspects of the Theory of Syntax/Remarks on Nominalizations view of Generative Grammar, to the Conditions on Transformations/GB/MP one. "

   my mind these are not examples of paradigm shifts, but are all very firmly within the generative paradigm.

            The later-minimalist-language-is-perfect-or-an-optimal- solution-to-interface-requirements view, however, has the flavour of being something genuinely different than what preceded it (the "biolinguists"'s claim to the contrary notwithstanding). Of course, it doesn't seem to have been brought about by any clear crisis of the earlier paradigm (but then, I haven't read much syntax in the last few years, so I may be off-base here).

          • daedalus2u says:

            An example that maybe linguists can better understand. How many languages can someone learn as first languages? What is it that distinguishes a first language from non-first languages?

            As Max Planck said

            A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.

            People can get stuck in their first language and never have the capacity to understand or speak or write a second language with the same fidelity and nuance as their first language. If your job was writing poetry in your first language, and then you had to start writing poetry in a non-first language, how long before that poetry would be at the same level as in your first language?

            There are some writers who do write great works in their non-first language, just as there are some scientists who forge new paradigms. I think that both tend to be exceptions.

        • daedalus2u says:

          If you have the "right" neural networks you will understand the idea. If you don't , you can't.

          Putting in the time is what causes your neural networks to self-modify until they can instantiate what ever idea you are trying to understand.

          Those neural networks can even instantiate doublethink, so I don't think there is a limit to what they can instantiate.

          Being incommensurate is about the path between the old neural network and the new neural network.

          • Avery Andrews says:

            "What is it that distinguishes a first language from non-first languages? "

            But, the paradigms are constituted by things people learn college or graduate school, long after the real critical periods are over. So I think other factors are needed to explain the de facto social reality of 'paradigms'.

            @Fred: I think one could make a case that the Aspects->Conditions/GB shift was about as big as the shift from some of Harris' or Lambek's work from early generativism was, and, in practice, the GB->MP shift was probably smaller. Note that virtually one of the main GB practitioners refused to get onto the MP bandwagon, while all of the 'alternative generative frameworks' that arose in the 7os were created and joined by people who had a problem with Conditions/GB (in my case, the downgrading of descriptive adequacy in the sense of capturing generalizations in an intelligable formalism).

          • Avery Andrews says:

            Gargh typoes:

            learn college -> learn in college

            virtually one -> virtually none

            This blog site needs an 'edit post' button

          • daedalus2u says:

            Avery, How do you know that there are not critical time periods for learning cognitive paradigms that are analogous to learning first language paradigms?

            If there are such time periods they would have to occur after the first language paradigm has stabilized language. Perhaps the cognitive structures that stabilize the first language paradigms are then used to stabilize cognitive, social and cultural paradigms during adolescence, and then perhaps stabilize gender, socialization and mating specific paradigms, and then still later stabilize parenting paradigms?

            There are a great many people that are stuck in pretty rigid social and cultural paradigms; religion, politics, sports team loyalty, monogamy, patriotism, xenophobia. Clearly there has to be physiology and neuroanatomy that produces and enforces such paradigm rigidity. I am not sure that there is much of a difference between loyalty to a religion and loyalty to Newtonian mechanics.

            Certainly scientists can get into arguments that are as heated as those of any religious or sports team or political zealots. Scientists and sports fans have ways of resolving those arguments, by experiment. Religious and political zealots do too, it is called war. Or in the title of the post, with ashtrays.

            In the comments on the last segment of the NYT column, there is a comment by Sarah Kuhn who says this is something completely uncharacteristic of her father and she disputes that it ever happened. I am inclined to believe her over the author of the columns.

          • Avery Andrews says:

            I don't *know* that there aren't critical periods for learning scholarly disciplines, but there is no real evidence that there are, and I think there are many simpler possible explanations for later life rigidity (say, in the 35-75 age-span, after that there is a much greater possibility of medical problems).

            One reason I don't sceptical of them is that I've had some excellent students who took up learning linguistics in their 50s after being housewives, school principals, etc., and various other random things, and they didn't seem to have any problems with it (better than most of the kids, in fact, due to superior motivation and discipline). Being established, in various ways, can for example simply reduce your motivation to study new things hard enough to learn them. The role of sociological factors in failure to learn is also a factor - we have a grad student whose topic involves rural Malaysian students of engineering who need to learn English but just won't.

            Also of course, the neural network idea ATM is not really a model, due to extreme lack of information about how they work in the areas we're discussing. It's comparable to the stuff I used to read as a kid in the 50s about the brain as a telephone switchboard.

          • daedalus2u says:

            I wasn't using “neural network” as a model, but as a description.

            When we see rigidity in many aspects of people's cognition, isn't that *evidence* of rigidity? Not everyone has rigid thinking, not everyone is rigid in the same ways, but rigidity of thinking is not rare. Presumably there is physiology that supports rigidity of thinking. My idea of parsimony is that the brain is unlikely to use a different mechanism for each different rigid style of thinking.

            You are correct, there aren't any good models of how the brain does what it does. We do know that whatever the brain does it does with neural networks. If we are ever going to understand what it is that the brain does, we are going to have to understand how it does it with neural networks.

            We do know there isn't a homunculus inside the brain making connections like a switchboard operator.

    • melodye says:

      Ha - yes. "God said earth wouldn't be destroyed by a flood." Oh good, so now we can just forget global warming because bible literalists haven't warmed to the idea of metaphor (and/or science, but that's really asking too much).

  • Jason says:

    Kuhn's ideas are essentially instantiations of Piaget's assimilation and accommodation (paradigm = schema; normal science = assimilation; revolution = accommodation). Kuhn acknowledges as much in the preface to the 2nd edition. If you read Piaget, I think he saves us from Kuhn's notion of incommensurability.

    • Avery Andrews says:

      I found Piaget to be tantalizingly almost but not quite intelligible. Not necessarily his fault.

  • Stan says:

    Wittgenstein preferred The Poker Proposition, supposedly. Though based on what I've read of the incident, the implement was merely brandished then dropped, not flung at Popper's head or anything like that.