[FRIAM] excess meaning alert? (was, Re: are we how we behave?)

[Nick Thompson]

Steve,

 

We were doing SO WELL until we got to … oh, see my “HORSEFEATHERS!” below. 

 

Nicholas S. Thompson

Emeritus Professor of Psychology and Biology

Clark University

 <http://home.earthlink.net/~nickthompson/naturaldesigns/> http://home.earthlink.net/~nickthompson/naturaldesigns/

 

From: Friam [mailto:friam-bounces at redfish.com] On Behalf Of Steven A Smith
Sent: Friday, March 29, 2019 9:39 AM
To: friam at redfish.com
Subject: Re: [FRIAM] excess meaning alert? (was, Re: are we how we behave?)

 

 

On 3/28/19 1:20 PM, Nick Thompson wrote:

Steve, ‘n all, 

 

Just to be cranky, I want to remind everybody that ALL language use, except perhaps tautological expressions, is metaphorical.

I ascribe to this idea as well, following Lakoff and Johnson in their 1980 _Metaphors we Live by_ .



  So then, the question is not, “Is this a metaphor”, but what kind of a metaphor is it and is it pernicious. 

I believe that ultimately conceptual metaphor is no more nor less than the intuitive application of a model, and as is often mentioned "all models are wrong, some are useful".    You use the term pernicious which suggests *harmful*, I presume either intentionally so or more from sloppiness or ignorance.



My own view is that in any “tense” conversation – one in which the parties feel the words really matter – it behooves a metaphor-user to define the limits of the metaphor. 

I agree that "tense" conversations are different than "casual" ones if that is your distinction.  Unfortunately, outside of Science/Engineering contexts, I find that "tense" conversations are at their root political or at least rhetorical.   One or both sides are really *serious* about being believed.   If not believed in fact ("I believe what you just said") then in principle ("I believe that you believe what you just said").

I think that political/rhetorical dialog would *benefit*  by careful disclosure of all metaphors being used, but one mode of such dialog is for one or both sides to attempt to interject equivocal meanings... to use a term (or in this case set of terms belonging to a metaphorical domain) to weave an *apparently* logical argument, which is only superficially logical but falls apart when the "correct" meaning of the term(s) are applied.   

So, for instance, much mischief has arisen in evolutionary biology from a failure of theorists to define the limits of their use of such metaphors as “natural selection” and “ adaptation”.  When limits are defined, the surplus meaning of a metaphor is separated into two parts, initially, that which the metaphor-user embraces and that which s/he disclaims.  The embraced part goes on to become the positive heuristic of the metaphor, the “wet edge” along which science develops. 

>From this line of discussion, I take you to be on the branch of the fault-tree I implied above as a Scientific dialog where *both* sides of the discussion are honestly trying to come to mutual understanding and perhaps advance understanding by combining differing perspectives on the same phenomena.



The disclaimed part, must be further divided into that which was legitimately [logically] disclaimed and that which was disclaimed fraudulently.  For instance, when sociobiologists use the notion of selfish gene, they may legitimately disclaim the idea that genes consciously choose between self-regarding and other-regarding options, but they cannot legitimately disclaim the idea that a gene has the power to make any choice but the self-regarding one.

When Dawkins coined "Selfish Gene",  I felt that the *value* of the metaphor invoked was in the challenge it presents:



  And that idea is patently false.  Genes do not make choices

Patently Genes do not make choices in the sense that we usually mean "make choices", yet the strong implication is that the phenomena functions *as if* they do, in "all other ways".   There may be (useful) hairsplitting between "all other ways" and "many other ways" which is an important aspect of analogical thinking.  



, they ARE choices and the choice is made at the level of the phenotype or at the level of the population, depending on how one thinks about the matter.  So the metaphor ‘selfish gene’ is pernicious in evolutionary biology, because it creates confusion on the very point that it purports to clarify – the level at which differential replication operates to generate long term phenotypic change in a population. 

I would challenge this as I think my verbage above outlines.   I do not believe that the metaphor *purports* to clarify what you say it does.  

[NST==> HORSEFEATHERS! One or two generations of sociobiologists were directed away from group level explanations by this pernicious metaphor.  <==nst] 

It *strives* to provide a cognitive shortcut and to establish a fairly strong metaphor which deserves careful dissection to understand the particulars of the *target domain*.   An important question in the target domain becomes "why does the shortcut of thinking of genes as selfish actually have some level of accuracy as a description of the phenomena when in fact the mechanisms involved do not support that directly?"

[NST==>I don’t think it does.  I think it’s a subtle and largely successful attempt to import Spenserian ideology in to evolutionary biology.  <==nst] 

For all I know, EB has entirely debunked the concept and there is NO utility in the idea of a "selfish gene"...  

Bruce Sherwood likes to make the point that the analogy of hydraulic systems for DC circuits is misleading.   I forget the specifics of where he shows that the analogy breaks down, but it is well below (or above?) the level of "normal" DC circuit understanding and manipulation.   For the kinds of problems I work with using DC circuits, a "battery" is a "tank of water at some height", the Voltage out of the battery is the water Pressure, the amount of Current is the Volume of water, a Diode is a one-way valve,  a resistor is any hydraulic element which conserves water but reduces pressure through what is nominally friction, etc.    As you point out, there is plenty of "excess meaning" around hydraulics as source domain, and "insufficient meaning" around DC circuits as target domain, and if one is to use the analogy effectively one must either understand those over/under mappings, or be operating within only the smaller apt-portion of the domains.   For example, I don't know what the equivalent of an anti-hammer stub (probably a little like a capacitor in parallel?) is but that is no longer describing a simple DC circuit. 

[NST==>I think I am back to heartily agreeing. <==nst] 

A farmer buying his first tractor may try to understand it using the source domain of "draft animal" and can't go particularly wrong by doing things like "giving it a rest off and on to let it cool down", "planning to feed it well before expecting it to work", "putting it away, out of the elements when not in use", etc.  your "excess meaning" would seem to be things like the farmer going out and trying to top off the fuel every day even when he was not using the tractor, or maybe taking it out for a spin every day to keep it exercised and accustomed to being driven.   The farmer *might* understand "changing the oil" and "cleaning the plugs" and "adjusting the points" vaguely like "deworming" and "cleaning the hooves" but the analogy is pretty wide of the mark beyond the simple idea that "things need attending to".

[NST==>OoooooH.  I like the above!  May I plaigiarise it some day?  Do you by any chance know Epamanondas from your childhood.  Very politically incorrect, now, I fear, but endlessly instructive on the perils of over using metaphors.  <==nst] 

 

PS – Is anybody on this list (among the handful that have gotten this far in this post) familiar with the work of Douglas Walton? 

I just took a look and his work does sound interesting (and relevant).



He seems perhaps to have written a lot about misunderstandings in AI systems … i.e., how does Siri know what we mean?  

By AI, it seems you mean (the subset of) Natural Language Understanding?

I am also reminded by reading the Wikipedia article on his work that I haven't responded to Glen's question about the "theorem dependency project".

I came to this work through my interest in abduction, which may be described as the process by which we identify (ascribe meaning to?) experiences.  Walton seems to suggest that you-guys are way ahead of the rest of us on the process of meaning ascription, and we all should go to school with you.  Please tell me where and when you offer the class.

I assume the "you-guys" referred to here are the hard core CS/Modeling folks (e.g. Glen, Marcus, Dave, ...).  I do think that the challenges of "explaining things to a machine" do require some rigor, as does formal mathematics and systems like the aforementioned "theorem dependency project".

- Steve

PS.  It has been noted that my long-winded explanation of my (poorly adhered to) typographical conventions for around "reserved terms" and the  like was perhaps defensive.  I didn't mean to sound defensive, I just wanted to be more precise and complete to (possibly) reduce misunderstandings.   I don't imagine many read the entireity of my missives, but as often as not,  when people do read and respond, I sense that some of my conventions are not recognized.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20190329/feb80644/attachment.html>