From Bill: Help! / linguistics

Re: From Bill: Help! / linguistics

It seems to me that there are huge problems in understanding how brains as they appear to work can produce anything like language as it appears to be, roughly as if you were to put a mess of marmalade and peanut butter into the microwave and get out a lego pirate ship. So people just struggle with this as best they can, with results that are fundamentally confusing and unsatisfactory. And then there are are the PCT specific problems, such as exactly what kind of error signal is the production of an utterance supposed to control, and how can you get from any kind of error signal to the production of something like:

“Sarah Palin enjoys an ability to connect with voters that cannot be taught”

(from somewhere on the interet, containing an entertaining and presumably unintended ambiguity)

So I have a few speculations of my own, but don’t know any way to get them above the level of what would be politely described as ‘just so stories’.

Avery Andrews

On Wed, 22 Jul 2009 10:42:57 -0600, Bill Powers powers_w@FRONTIER.NET wrote: >[From Bill Powers (2009.07.22.1027 MDT)] > > >How about introducing yourself a bit, since I don’t recall hearing >from you before now? I hope that’s not just senility setting in. > >Best, > >Bill P. >========================================================================= No, it’s rather my lack of manners not introducing myself… … I’m just an interested bystander who stumbled over CSGnet a while ago and became somehow hooked. I did a Master in linguistics a few years ago in germany (where I was born and live). During my studies I became very frustrated with my subject. Everywhere I looked, e.g. psycholinguistics and neurolinguistics, models where based on the input-output view and/or the “brain-as-a-computation-device” paradigm - which both never made sense to me. Take e.g. Alan Baddeley’s model of “working memory” which is still very popular among psychologists. The model consists of buffers (for temporary storage), modules and it assumes the transmission of information between these modules. At my university students still have to learn Chomsky and similar “cognitive” approaches on language. More recent theories of language are still understood within the same traditional paradigm. There seems to be no real attempt to understand language on a biological level. Of course, there are many so called “biologically inspired” theories, and of course, most of them claim to be based on “facts” of how the brain works. But unfortunately the brain is no computation device. If this were true, you PCTers would have easily transmitted your information/knowledge into the heads of these scientists out there. (Of course the real problem is related to the phenomenon of reorganisation … which organisms usually prefer to resist). Anyway, I enjoy following your discussions. The PCT-understanding of ‘behavior’ is a refreshing alternative to these biologically unplausible models, which usually impy a kind of homunculus (e.g. “central executive”). I don’t see any way how to understand behavior other than as the control of perception. However, what I’m missing on CSGnet is the attempt to link PCT more to structures and properties of the nervous system. And I would be interested to hear the PCT view on language at the perceptual and the behavioral structural, not only the functional, level. I have some half baked ideas… which I will be happy to share, provided that you bear my weak command of english. Regards, Ulrich

Back to: Top of message | Previous page | Main CSGNET page

Lists

Hi Avery,

Studies like the one in this news article
(http://news.yahoo.com/s/ap/us_med_healthbeat_bilingual_tots) show
that the language references are developed very early in life. I have
a five month old daughter, and daily I watch the emergence of new
references (not just regarding language). I speculate also that the
intricacy of the development of language references is why humans have
evolved to be so helpless in the first year. This helplessness
encourages human infants to control their perceptions through X. I
cannot think of how to discuss/describe X that Bill or Mark would
accept. So, I will let you try if you like.

Regards,
Shannon Williams

···

On Fri, Jul 24, 2009 at 8:20 PM, Avery Andrews<Avery.Andrews@anu.edu.au> wrote:

It seems to me that there are huge problems in understanding how brains as
they appear to work can produce anything like language as it appears to be,
roughly as if you were to put a mess of marmalade and peanut butter into the
microwave and get out a lego pirate ship.� So people just struggle with this
as best they can, with results that are fundamentally confusing and
unsatisfactory.� And then there are are the PCT specific problems, such as
exactly what kind of error signal is the production of an utterance supposed
to control, and how can you get from any kind of error signal to the
production of something like:

�"Sarah Palin enjoys an ability to connect with voters that cannot be
taught"

(from somewhere on the interet, containing an entertaining and presumably
unintended ambiguity)

So I have a few speculations of my own, but don't know any way to get them
above the level of what would be politely described as 'just so stories'.

�Avery Andrews

On Wed, 22 Jul 2009 10:42:57 -0600, Bill Powers <powers_w@FRONTIER.NET>
wrote: >[From Bill Powers (2009.07.22.1027 MDT)] > > >How about introducing
yourself a bit, since I don't recall hearing >from you before now? I hope
that's not just senility setting in. > >Best, > >Bill P.

=========================================================================

No, it's rather my lack of manners not introducing myself... ... I'm just an
interested bystander who stumbled over CSGnet a while ago and became somehow
hooked. I did a Master in linguistics a few years ago in germany (where I
was born and live). During my studies I became very frustrated with my
subject. Everywhere I looked, e.g. psycholinguistics and neurolinguistics,
models where based on the input-output view and/or the
"brain-as-a-computation-device" paradigm - which both never made sense to
me. Take e.g. Alan Baddeley's model of "working memory" which is still very
popular among psychologists. The model consists of buffers (for temporary
storage), modules and it assumes the transmission of information between
these modules. At my university students still have to learn Chomsky and
similar "cognitive" approaches on language. More recent theories of language
are still understood within the same traditional paradigm. There seems to be
no real attempt to understand language on a biological level. Of course,
there are many so called "biologically inspired" theories, and of course,
most of them claim to be based on "facts" of how the brain works. But
unfortunately the brain is no computation device. If this were true, you
PCTers would have easily transmitted your information/knowledge into the
heads of these scientists out there. (Of course the real problem is related
to the phenomenon of reorganisation ... which organisms usually prefer to
resist). Anyway, I enjoy following your discussions. The PCT-understanding
of 'behavior' is a refreshing alternative to these biologically unplausible
models, which usually impy a kind of homunculus (e.g. "central executive").
I don't see any way how to understand behavior other than as the control of
perception. However, what I'm missing on CSGnet is the attempt to link PCT
more to structures and properties of the nervous system. And I would be
interested to hear the PCT view on language at the perceptual and the
behavioral structural, not only the functional, level. I have some half
baked ideas... which I will be happy to share, provided that you bear my
weak command of english. Regards, Ulrich

Back to: Top of message | Previous page | Main CSGNET page

Lists

[From Fred Nickols (2009.07.25.0906 EDT)]
  

From: Avery Andrews <Avery.Andrews@ANU.EDU.AU>

It seems to me that there are huge problems in understanding how brains as they
appear to work can produce anything like language as it appears to be, roughly
as if you were to put a mess of marmalade and peanut butter into the microwave
and get out a lego pirate ship. So people just struggle with this as best they
can, with results that are fundamentally confusing and unsatisfactory. And then
there are are the PCT specific problems, such as exactly what kind of error
signal is the production of an utterance supposed to control, and how can you
get from any kind of error signal to the production of something like:

"Sarah Palin enjoys an ability to connect with voters that cannot be taught"

(from somewhere on the interet, containing an entertaining and presumably
unintended ambiguity)

So I have a few speculations of my own, but don't know any way to get them above
the level of what would be politely described as 'just so stories'.

It seems to me that almost all of us are born with the capability for using language but that which language we use and how we use it is more a function of learning to do so in a particular setting. We aren’t born with the reference signals for uttering a particular word, we acquire them. We hear others speaking, we can recall how that sounded, we can make sounds ourselves, comparing them with what we recall, and adjust our utterances until we are making similar sounds. That’s basically simple imitation.

But, right now, if PCT is correct, there is a set of reference signals at work as I formulate this sentence. My fingers on the keyboard lag behind my sub-vocal speech and what my eyes perceive can be matched against my just past and easily recalled sub-vocal utterances. (I can also return to this text and alter it so that what appears in print more closely aligns with my reference signals for that kind of thing.)
  
The reference signals that allow me to compare printed words with uttered words were established long ago and not all of them correctly, by the way; I’ve had to modify some over the years. For example, for many years, I pronounced the printed word “misled�? as “muzzled.�? Don’t have a clue as to how that happened, but it did and it was only corrected a few years ago.

So, one part of the issue is mastering a vocabulary. Another part consists of speaking and writing (two different things). Suppose I say, “I’m going to go antiquing today.�? Do I have a reference signal for that particular utterance? My answer would be yes and no. The Yes answer is that I do have reference signals that make my utterances conform to certain standards but at varying levels of detail. One of them is probably something that results in utterances that approximate correct grammar. Ditto for my written words. But, do I have a reference signal against which this piece I’ve just written (and which I uttered sub-vocally) can be compared line-by-line, word-by-word, so that a judgment could be made that the end product is consistent with a pre-existing reference signal? I don’t think so.

In terms of a single word (let’s use “terminological�? as an example), I think that the sounds we utter are adjusted as we utter them so as to match the reference signal we have for how that word should sound. “Terminological�? has several syllables so we make several sounds in the course of uttering it. In terms of which words we utter in which sequences with which emphasis, I think the reference signals exists at a higher level, possibly in the form of “Does what I’ve just said or written make sense?�? and “Does it convey the meaning I want to convey?�? Or, perhaps, “Was what I just said or wrote responsive to the question or the matter at hand?�? I can make those kinds of judgments, too, and they are much more conscious and deliberate than merely uttering this or that word correctly (or consistent with what I believe to be correct).

The bottom-line for me is that many reference signals emerge, formulated in the course of interacting with the world around me. When it comes to language, some reference signals (e.g., pronunciation and grammar and word meanings/definitions) are relatively fixed and pre-existing. Other reference signals (e.g., what to say and how to phrase it) are formulated in real time. As I understand PCT, a closed loop is “live�? all the way around � reference signals and perceptions interact continuously aand where you choose to start in examining that kind of loop is somewhat arbitrary.

In any event, speaking and writing are clearly dynamic occurrences, not step-wise, pre-programmed production routines that have been stored somewhere waiting to be activated. Or, at least, so I believe.

Egad! I wonder what set me off this morning! Hmm. I wonder if this is a case of positive feedback and a runaway system that produces nothing except babbling?

···

-------------- Original message ----------------------

--

Regards,

Fred Nickols
Managing Partner
Distance Consulting, LLC
nickols@att.net
www.nickols.us