[From Bill Powers (2004.06.01.0814 MDT)]
Bruce Nevin (05.31.2004 22:50 EDT) --
Bruce, I can see that my clumsy efforts to get across my point of view are
arousing more irritation than interest, so I'll give it a rest. It's not as
though I have everything worked out to my satisfaction, not to mention
anyone else's.
Best,
Bill P.
From [Marc Abrams (2004.06.01.1105)]
[From Bill Powers (2004.06.01.0814 MDT)]
Bruce, I can see that my clumsy efforts to get across my
point of view are arousing more irritation than interest, so...
[From Bill Powers (2004.06.01.0757 MDT)]
As I am unable to communicate my ideas about how to start modeling
economic systems or anything else (the
defenses are too much for me to overcome)...
Ironic, I feel he same way you do about you. I wonder if it's a control
thing?
Marc
Considering how often throughout history even intelligent people have
been proved to be wrong, it is amazing that there are still people who
are convinced that the only reason anyone could possibly say something
different from what they believe is stupidity or dishonesty.
Being smart is what keeps some people from being intelligent.
Thomas Sowell
Don't argue with an idiot; people watching may not be able to tell the
difference.
Anon
I don't approve of political jokes. I've seen too many of them get
elected
Anon
[From Bruce Nevin (06.02.2004 22:14 EDT)]
Bill Powers (2004.06.01.0814 MDT)–
Bruce, I can see that my clumsy efforts to get
across my point of view arearousing more irritation than interest, so I’ll give it a rest. It’s not
asthough I have everything worked out to my satisfaction, not to
mentionanyone else’s.
Huh? I thought we were pretty close to agreement. Nor did I think that I
was being particularly testy. Well, granted, I wrote it tired and in a
hurry, and without my usual practice of waiting and rewriting.
Let’s see if I understand the point of view that you are trying to get
across. The point of view that you have articulated (as understand it, of
course, which goes without saying, but it seems that I need to say it)
does explain how social phenomena like the rings and arcs of the CROWD
program can emerge as side effects of individuals’ control in an
environment where a number of like individuals are controlling the same
variables in like ways. These phenomena are uncontrolled side effects.
They are not themselves perceived by the agents involved (in the CROWD
program), or if they are perceived (by humans) the perception is
irrelevant to the emergence of the phenomena. All that is required in the
CROWD program is that each of the agents controls proximity to the goal
with one reference value and proximity to other objects (including other
agents) with another. You have suggested that all cultural and social
phenomena could be emergent side effects of control just like these.
The point of view that you have articulated does not consider how the
agents came to distinguish between the goal and the obstacles, nor how
any given agent came to control with those reference values for approach
to the goal and avoidance of the obstacles, nor how it happens that they
all make the same perceptual distinction between goal and other objects
with the same reference values. These are not insurmountable problems,
but they are worth mentioning.
Insofar as all control systems pursue goals and avoid obstacles, these
phenomena are universal. They necessarily emerge whenever a sufficient
number of autonomous control systems pursue the same goal.
Cultural phenomena are not universal. They differ from one community to
another. For example, words and sentences in the Farsi language are
absent from our prior exchanges on CSGnet but commonplace in verbal
exchanges in Tehran.
There is strong evidence that humans assert group membership and identify
others as non-members (or as members of specific groups) by controlling
perceptions of cultural phenomena. It appears that cultural
differentiation is in large measure an effect of this process. This is
obviously circular, but that shouldn’t trouble us. We’re accustomed to
circular causation.
I have not been arguing for the ultimate really real Reality of cultural
phenomena, only for the much weaker position that there is something
going on that PCT must account for, and which the point of view that you
have advanced (as I understand it) does not allow for consideration. And
I have tried to identify some characteristics of these phenomena. I have
done this partly just to distinguish non-universal cultural phenomena
from universal byproducts of control such as those demonstrated by the
CROWD program. In the face of claims that the CROWD program indicates a
paradigm of research that can be expected to account for all social
phenomena, it has seemed necessary to argue for the mere existence of
social phenomena that are culturally differentiated, and to investigate
how a simulation would have to be organized in order to model such
phenomena. Here are some characteristics that seem to me to be
essential.
At 08:18 AM 6/1/2004 -0600, Bill Powers wrote:
by means of
by means of
These seem to me to be essential for a model of control systems
developing and enacting culture.
The last two characteristics-- (4) historical contingency &
cumulative development, and (5) divergence into distinct cultures – do
not tell us how to construct a model, but they can perhaps help to tell
us if a model is behaving appropriately.
People may be deluded in making generalizations about others and
perceiving those generalizations as social realities. The peculiar thing
is that by doing so, collectively and reciprocally, the generalizations
become accurate perceptions of what one another is doing. The effect is
to enhance stability and predictability in social interactions by means
of mutually confirming expectations. It has all the speciousness of a
self-fulfilling prophesy, but it’s a useful speciousness.
/Bruce
Nevin
[From Bill Powers (2004.06.03.0130 MDT)]
Bruce Nevin (06.02.2004 22:14 EDT)--
Bruce, as I see it our differences revolve not around the question of
whether social phenomena exist, but where they come from. I think that if
we could somehow model a whole human being at all the levels, including the
capacity to reorganize and properties of the environment, nothing else
would be required but duplicating that model as many times as necessary to
recreate all the social phenomena ever observed -- or at least, phenomena
of the same kinds if not the same in every detail. I think that is a way of
saying that social phenomena are emergent, and do not stem from any source
other than the active agents involved.
It has seemed to me that you are claiming otherwise -- that there are
social laws and social forces that come from some source other than the
individuals who make up the society. I see similar implications in Martin
Taylor's and Bill Williams' discussions. Perhaps I am simply misreading; if
so, it's about time to straighten out that problem with my dyslexia.
Best,
Bill P.
[From Bruce Nevin (06.03.2004 09:02 EDT)]
Bill Powers (2004.06.03.0130 MDT) --
Bruce, as I see it our differences revolve not around the question of
whether social phenomena exist, but where they come from. I think that if
we could somehow model a whole human being at all the levels, including the
capacity to reorganize and properties of the environment, nothing else
would be required but duplicating that model as many times as necessary to
recreate all the social phenomena ever observed -- or at least, phenomena
of the same kinds if not the same in every detail. I think that is a way of
saying that social phenomena are emergent, and do not stem from any source
other than the active agents involved.It has seemed to me that you are claiming otherwise -- that there are
social laws and social forces that come from some source other than the
individuals who make up the society. I see similar implications in Martin
Taylor's and Bill Williams' discussions. Perhaps I am simply misreading; if
so, it's about time to straighten out that problem with my dyslexia.
Bill,
You are reading into it things that I am not saying. I have identified
specific kinds of variables such that, if control systems are controlling
them, cultural phenomena will emerge. These are
1. Controlling aspects of behavioral outputs ("manner of controlling").
This can be used as means of
2. Controlling social affiliation. (Crudely: who are my allies, who are not.)
And this can be used as means of
3. Controlling a perception of self in context of others.
/Bruce Nevin
At 01:38 AM 6/3/2004 -0600, Bill Powers wrote:
From Bill Williams 3 May 2004 9:40 AM CST]
I am attempting to catch up to posts in sequence, but I would like to say that I am in agreement with with Bill Powers when he says,
[From Bill Powers (2004.06.03.0130 MDT)]
Bruce, as I see it our differences revolve not around the question of
whether social phenomena exist, but where they come from. I think that if
we could somehow model a whole human being at all the levels, including the
capacity to reorganize and properties of the environment, nothing else
would be required but duplicating that model as many times as necessary to
recreate all the social phenomena ever observed -- or at least, phenomena
of the same kinds if not the same in every detail. I think that is a way of
saying that social phenomena are emergent, and do not stem from any source
other than the active agents involved.
I would regard any other type explaination as magical.
It has seemed to me that you are claiming otherwise -- that there are
social laws and social forces that come from some source other than the
individuals who make up the society.
I recognize that this is what you think, but I can not account for why you
would think this.
I see similar implications in Martin Taylor's and Bill Williams' discussions.
Perhaps I am simply misreading; if so, it's about time to straighten out >that problem with my dyslexia.
I think habitual associations is a more likely source. But, I wold agree that the issue is a fundamentally important one. Many and perhaps most people who want to concern questions of social theory make precisely the mistake you are so admantly, and correctly opposed to-- the Durkheiemian social forces doctrine. I don't perceive that Martian Taylor, Bruce Nevin nor myself are making that mistake. With some differences in expression, I see myself in agrement with martin and bruce.
Bill Williams
[From Bill Powers (2004.06.03.0747 MDT)]
Bruce Nevin (06.03.2004 09:02 EDT)--
You are reading into it things that I am not saying. I have identified
specific kinds of variables such that, if control systems are controlling
them, cultural phenomena will emerge. These are1. Controlling aspects of behavioral outputs ("manner of controlling").
This can be used as means of
2. Controlling social affiliation. (Crudely: who are my allies, who are not.)
And this can be used as means of
3. Controlling a perception of self in context of others.
This must be a very specialized definition of "self." To me, "myself"
simply means me, this person, as opposed to someone else, that person. I
don't see that personal affiliations are particularly relevant to who or
what I am. I don't "affiliate" with myself, but with others, and others are
different from me. That's how I know they are "others.". But I recognize
that some people do think affiliations are relevant in defining themselves,
which of course makes them relevant to those people.
When you say that controlling aspects of behavioral outputs (meaning,
controlling perceptions of them) can be used as a means of controlling
social affiliation, you're automatically selecting just those persons for
whom this is a goal. And when you say that social affiliations are
controlled as a means of controlling a perception of self, you further
restrict the applicability of these ideas to individuals who define
themselves primarily in terms of how they relate to others. So how big is
this final group? Are you saying that this specific hierarchy fits everyone?
Best,
Bill P.
[Martin Taylor 2004.06.03.1049]
[From Bill Powers (2004.06.03.0130 MDT)]
Bruce Nevin (06.02.2004 22:14 EDT)--
... I think that if
we could somehow model a whole human being at all the levels, including the
capacity to reorganize and properties of the environment, nothing else
would be required but duplicating that model as many times as necessary to
recreate all the social phenomena ever observed -- or at least, phenomena
of the same kinds if not the same in every detail. I think that is a way of
saying that social phenomena are emergent, and do not stem from any source
other than the active agents involved.It has seemed to me that you are claiming otherwise -- that there are
social laws and social forces that come from some source other than the
individuals who make up the society. I see similar implications in Martin
Taylor's and Bill Williams' discussions. Perhaps I am simply misreading;
I suspect that you are. I can vouch for the fact that you are in my
case. I completely agree with your first paragraph, and therefore not
with your second.
In one of your messages, you echoed (as if contradicting me) a claim
I made, that a cultural artifact had the same reality status as a
rock. No more, no less. Both are constructed perceptions.
Bruce and I have a slight divergence of emphasis, in that Bruce
emphasises the use of mimicry: our ability to perceive our outputs
and bring those perceptions to reference values that correspond to
our perceptions of other people's actions. I emphasise the
"theoretical fact" that reorganization from infancy onward, without
the need to perceive our own output, will lead us to develop actions
that effectively allow us to control our perceptions. In a social
context, that will happen optimally when our actions are sufficiently
like those of others in the same cultural group to influence them to
do what we want to have done.
I'm not discounting mimicry, and I doubt that Bruce is discounting
normal reorganization. It's just a different emphasis.
My own point is simply that one can kick against a cultural artifact
and get a stubbed toe just as readily as one can do so with a rock.
They are kicks and stubbed toes at different levels of perceptual
abstraction, but each can hurt quite badly!
Martin
[From Bill Powers (2004.06.03.0914 MDT)]
Bill Williams 3 May 2004 9:40 AM CST--
...I would agree that the issue is a fundamentally important one. Many and
perhaps most people who want to concern questions of social theory make
precisely the mistake you are so admantly, and correctly opposed to-- the
Durkheiemian social forces doctrine. I don't perceive that Martian
Taylor, Bruce Nevin nor myself are making that mistake. With some
differences in expression, I see myself in agrement with martin and bruce.
Then perhaps you can explain to me what you mean when you say there is no
such thing as an individual. It seems to me that if we eliminate the
Durkheim forces, individuals are all that exist. Is this another of those
specialized uses of terminology?
Best,
Bill P.
[From Bill Powers (2004.06.03.0931 MDT)]
Martin Taylor 2004.06.03.1049 --
In one of your messages, you echoed (as if contradicting me) a claim
I made, that a cultural artifact had the same reality status as a
rock. No more, no less. Both are constructed perceptions.
You failed to say what the reality status of the rock was. I thought you
were saying it was real, therefore so is the cultural artifact.
Bruce and I have a slight divergence of emphasis, in that Bruce
emphasises the use of mimicry: our ability to perceive our outputs
and bring those perceptions to reference values that correspond to
our perceptions of other people's actions.
But when we mimic things like gestures, we don't recreate what we see. We
do things which we imagine would look like what the other is doing from an
outside point of view. When I make a face, I have no idea how it looks
unless I practice in front of a mirror (discounting the reversal). People
even find their own voices strange when listening to a hi-fi recording.
Whatever mimicry is, it's not just reproducing a remembered perception.
I emphasise the "theoretical fact" that reorganization from infancy
onward, without the need to perceive our own output, will lead us to
develop actions
that effectively allow us to control our perceptions. In a social
context, that will happen optimally when our actions are sufficiently
like those of others in the same cultural group to influence them to
do what we want to have done.
I agree, of course, within the loose construction of cause and effect. But
I don't think similarity is necessarily the key. When I want something I
don't say "You're welcome," but "please." The _other_ person says "You're
welcome." I think perception of the social variables is the key. We vary
our behavior in relation to other people to control social variables,
learning what works as we go.
My own point is simply that one can kick against a cultural artifact
and get a stubbed toe just as readily as one can do so with a rock.
They are kicks and stubbed toes at different levels of perceptual
abstraction, but each can hurt quite badly!
Yes. Within the world of mutual agreement, as near as we can ascertain what
that is, we can agree on that. My reservation is that we do not have nearly
as much reliable knowlege about how "people" behave in social situations as
one would guess from reading the literature. And very few of the things on
which we try to reach agreement are as simple as a rock. Don't forget that
average correlation of 0.26.
Best,
Bill P.
From[Bill Williams 3 May 2004 11:00 AM CST]
[Martin Taylor 2004.06.03.1049]
I'm not discounting mimicry, and I doubt that Bruce is discounting
normal reorganization. It's just a different emphasis.
I found your restatement of Bruce Nevin's argument and a concise
restatement of your own position helpful. I think that I agree
with both of your positions. And, I am in agreement with the following.
My own point is simply that one can kick against a cultural artifact
and get a stubbed toe just as readily as one can do so with a rock.
They are kicks and stubbed toes at different levels of perceptual
abstraction, but each can hurt quite badly!
However, we may not be quite over the hump in communicating that the
"cultural artifact" isn't one more in a series of magical attributions
to a "group" of properties that I would argue are only properies of a
in my terms "specimen."
Bill Williams
From[Bill Williams 3 May 2004 11:20 AM CST]
[From Bill Powers (2004.06.03.0914 MDT)]
Bill Williams 3 May 2004 9:40 AM CST--
...I would agree that the issue is a fundamentally important one. Many and
perhaps most people who want to concern questions of social theory make
precisely the mistake you are so admantly, and correctly opposed to-- the
Durkheiemian social forces doctrine. I don't perceive that Martian
Taylor, Bruce Nevin nor myself are making that mistake. With some
differences in expression, I see myself in agrement with martin and bruce.
Then perhaps you can explain to me what you mean when you say there is no
such thing as an individual.
As I understand it you didn't find anything to which you objected in my intitial post in this thread. I think the lattic program might serve as a basis for a second step in my argument building upon that first post.
In a previous post today, I explained how in the lattice program at each interation each of the squad members adjusted their reference position to a point on the field which would be equally distant from each of their fellow squad members. The do so only because they have choosen to be a member of the group and have agreed to adopt this norm. Now their behavior, it is still autonomous in the sense that they can still choose to repudiate their earlier choice is they wish, has as a result of their choice a social aspect. And, we can see the results of this when the group forms up. There is no "group" that does the forming up-- it is still each "individual" (I use quotes to indicte the enourmous baggage that this term carries) or specimen does everything that gets done. In this case by attempting to maintain an interval left-right front-back that is equal. So, at each step a new reference level is adopted-- the perception of distance that will make the intervals equal. Then the specimen attempts to move so as to actually make the intervals equal. The result is a coordinated pattern of specimens on a parade groud. And, depending upon how the information upon the specimen's is transmitted the behavior of two groups can be quite different even though the speciemens in the two groups are all the same.
The assertion "there is no such thing as an indivigual" is taken from H.H. Cooley who was one of Dewey's students. The full quote is ( approximately )
"There is no such thing as an individual and there is in the same sense no such thing as a collective." Updated, I would say that "All of the agency is in the specimen, but specimens make different choices depending upon which group or groups they agree to share an environment with." Such a formulation would place all of the initiative within the discrete biological specimen. It would also recognize that an agreement of such a discrete specimen will make different choice depending upon which group the specimen is a member of. So, the group is important but not in the way social theorists have usually thought of the group having an importance.
I hope the Lattice program provides a context for coming to an agreement as to what is involved.
It seems to me that if we eliminate the
Durkheim forces, individuals are all that exist. Is this another of those
specialized uses of terminology?
I don't think so. But, lets see if starting from the standpoint of what is going on in the lattic program provides a context for answering the question.
Bill Williams
[From Bill Powers (2004.06.03.1037 MDT)]
Bill Williams 3 May 2004 11:20 AM CST--
In a previous post today, I explained how in the lattice program at each
interation each of the squad members adjusted their reference position to
a point on the field which would be equally distant from each of their
fellow squad members.
I'd forgotten that rationale behind the program, It seems to me that you
would need to introduce the level at which this "equidistant" relationship
is perceived and controlled by varying the reference for position in X and Y.
I would start by defining the controlled perception. We could say that the
control systems can perceive (at a lower level) the distance to neighboring
points, each specimen perceiving in X and Y independently. The particular
points being observed depend on where in the array the specimen is; for
those on the boundaries, there would one boundary point and one interior
neighbor; for the rest, there would be two interior neighbors. This is
following your computations closely.
From these two perceptions could be constructed a perception of
"difference in distance to the two points." The reference level for this
difference would be zero, indicating no difference and hence equal
distance. The error signal would be used to adjust the reference position
for the system in question (or you could just convert the error into a
change of position, skipping the model of the lower-level system).
The do so only because they have choosen to be a member of the group and
have agreed to adopt this norm. Now their behavior, it is still
autonomous in the sense that they can still choose to repudiate their
earlier choice is they wish, has as a result of their choice a social
aspect. And, we can see the results of this when the group forms
up. There is no "group" that does the forming up-- it is still each
"individual" (I use quotes to indicte the enourmous baggage that this
term carries) or specimen does everything that gets done. In this case
by attempting to maintain an interval left-right front-back that is
equal. So, at each step a new reference level is adopted-- the perception
of distance that will make the intervals equal.
Ah,that explains why you change the reference signal to match the
perceptual signal on each step. But try it my way -- I think you'll see
that it makes more sense and works at least as well. The perceptual signal
after each step is not really the perception of distance that will make the
intervals equal -- not until they are actually equal. The principle of
control to use here is comparison of the perceived state with the reference
state, the reference state defining what the perception will be when the
error is zero, and the perception defining what the current state of
affairs is -- not the zero-error state.
The assertion "there is no such thing as an indivigual" is taken from H.H.
Cooley who was one of Dewey's students. The full quote is ( approximately )
"There is no such thing as an individual and there is in the same sense no
such thing as a collective." Updated, I would say that "All of the agency
is in the specimen, but specimens make different choices depending upon
which group or groups they agree to share an environment with."
I think that can be simplified to "specimens control their perceptions by
acting on the world they happen to be in." This remains true whether the
environment contains other active specimens or only passive physical elements.
Such a formulation would place all of the initiative within the discrete
biological specimen.
Yes, that is my view, too.
It would also recognize that an agreement of such a discrete specimen will
make different choice depending upon which group the specimen is a member of.
Just as it will make different choices depending on what environment it is
in, whether or not the environment contains other nearby specimens.
So, the group is important but not in the way social theorists have
usually thought of the group having an importance.
Yes. And the environment is important, but not in the way behaviorists have
thought of environments as having importance.
I don't think so. But, lets see if starting from the standpoint of what is
going on in the lattic program provides a context for answering the question.
That's fine. Why not rewrite the lattice program as I have suggested, and
then we can see the effect of varying the specimens' reference level for
difference in distance, and also varying the perception of distance, and of
introducing random variations in the reference levels of perceptual
functions. If we leave the lower-order position control in place, we can
show a two-level system at work, and by adding a third level, we can
introduce a place to speculate about why each actor has chosen the
reference level for spacing that is observed -- that is, what higher-level
perception is controlled by maintaining equal rather than unequal spacing..
As it is, the lattice program doesn't really have a control system for
controlling the most critical variable, equality of spacing. The control
systems control only position.
Best,
Bill P.
P.S. I'm off to a pre-operative exam for the rest of the day. Tomorrow I
head for Phoenix to spend the weekend and Monday morning talking with Ed
Ford's people, returning Tuesday just in time for the actual Wednesday
operation on the right eye. Left eye two weeks later. I'll try to stay in
touch but don't interpret silence as anything but being busy.
[From Rick Marken (2004.05.1030)]
Bill Williams (3 May 2004 11:20 AM CST) --
Bill Powers:
It seems to me that if we eliminate the Durkheim forces,
individuals are all that exist.
I don't think so. But, lets see if starting from the standpoint of what is
going on in the lattic program provides a context for answering the question.
I agree. Let's use the lattice program as a context for understanding what
this discussion is about. So far, the whole discussion about "Social
reality" sounds like much ado about nothing. Or, if not nothing, then about
explanatory capabilities that already exist in PCT. For example, Bruce
Nevin (06.03.2004 09:02 EDT) says:
I have identified specific kinds of variables such that, if control systems
are controlling them, cultural phenomena will emerge. These are1. Controlling aspects of behavioral outputs ("manner of controlling").
This can be used as means of
2. Controlling social affiliation. (Crudely: who are my allies, who are not.)
And this can be used as means of
3. Controlling a perception of self in context of others.
But this is all part of PCT already. Control of perceptions of style (the
manner of their own controlling) involves control of transitions, sequences
and events; control of perceptions of social affiliation involves control of
relationships and system concepts; controlling a perception of "self"
probably involves control of system concepts.
It would help me understand what this discussion is about -- and what the
disagreements actually are -- if we could make it more concrete. If you
think your lattice program is relevant to this topic then, by all means,
let's take a close look at it. I'd like to know what cultural phenomenon is
being modeled in the program and how does the model works.
Best
Rick
--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400
From[Bill Williams 3 May 2004 12:40 PM CST]
I think using the context of the Lattice program narrows down the
area of disagrement-- which there may still be one or two.
I'll ignore the areas of agrement and focus on what may be an area
of disagreement.
[From Bill Powers (2004.06.03.1037 MDT)]
Bill Williams 3 May 2004 11:20 AM CST--
The assertion "there is no such thing as an indivigual" is taken from H.H.
Cooley who was one of Dewey's students. The full quote is ( approximately )
"There is no such thing as an individual and there is in the same sense no
such thing as a collective." Updated, I would say that "All of the agency
is in the specimen, but specimens make different choices depending upon
which group or groups they agree to share an environment with."
I think that can be simplified to "specimens control their perceptions by
acting on the world they happen to be in." This remains true whether the
environment contains other active specimens or only passive physical >elements.
The difference between us may be in how this "enviornment" is defined. I am not sure I understand on first reading your suggestion as to how to modify the Lattice program-- to do it as you say, "your way." I would, however, agree that the program as it is now written depends upon an implicit understanding of the process the program is, in part, modeling.
I am not sure whether or not I can translate the elements which are implict into code. Writting the lattice program took me about three years. And, while the result is modestly complex the basic idea involved seems rather simple. It even appeared to me rather simple to start with, yet getting to the later versions required an enourmous effort. The problem you are suggesting, at least initially seems to me to involve elements that are much more complex than lattice program. So, to start with I don't have the slighest idea how to go about doining what you suggest. Not, that I don't see what you suggest as a logical next step.
As I perceive it the problem involved in improving the Lattice program is one of how to model the relationship of agreement between two specimens. But this may not be consistent with what you are suggesting.
Bill Williams
From[Bill Williams 3 June 2004 1:00 PM CST]
[From Rick Marken (2004.05.1030)]
Bill Williams (3 May 2004 11:20 AM CST) --
It would help me understand what this discussion is about -- and what the
disagreements actually are -- if we could make it more concrete. If you
think your lattice program is relevant to this topic then, by all means,
let's take a close look at it. I'd like to know what cultural phenomenon is
being modeled in the program and how does the model works.
Because of sequencing your post came in following a discussion between Bill Powers and I that adresses, at least in part, your suggestion.
As I understand it (and subject to correction) , Bill Powers doesn't object to how the existing lattice program operates. However, he argues that there is a lot that is implicit in the program. To make sense of it, I rely on the story about a group of, I'll call them "specimens" , who have agreed to form up. To do so they know five things: who is to their left, their right, their front and their rear, and that they are supposed to position themselves so that they are equally distant from these four specimens who are members of the group. For the people ( specimens ) on the edge of the formation there are chalkmarks that define the boundaries of the formation.
So, every specimen has learned how to form up. The specimen knws who, or what to form up on, and the principle involved in forming up-- make your left, right, front back interval equal. Somebody laid down the chalk marks laid out in a rectangular pattern.
To begin with everyone is standing in the middle of the field. This means that the everyone has equal intervals-- except for the specimens who are supposed to form up on the chalk marks. They have an enormous error, so they begin moving so as to make there position equally distant from their assigned left,right,front,back squad members. So, they start moving to a position halfway between the chalk mark reference and the center of the field where everyone else is standing. This creates an error for the specimens next to them, and they start moving. If the cooficients are all such that the control loops within the idividuals are stable the movements-- each one as the result of errors within the specimens will eventually result in a regular formation.
regular in the sense that all the intervals are equal.
Now, the diffrence between the information flows between the A and the B version of the program comes into play. When the group has formed up and stablized so that each specimen has an equal interval left,righ,front,back then the two different patterns in the flow of information can be tested.
In the A version the response of the formation to a disturbance is robust and after a disturbance -- a disturbanced introduced by pulling a specimen out of position-- the formation reforms with out extensive oscilations, or secondary disturbances. In the A version the information for each specimen is updated on every cycle so their is a short loop involved. In the B version, however, the initial disturbance travels across the formation, and diverges as it travels because of a delay in the information upon which each of the specimens depends.
I drew a "moral implication" from this difference that a society in which everyone has a quick and ready access to information would be less subject to disruption than a society in which the flow of information is restricted and delayed as it travels through the "group."
At this point I don't know idea how many people have seen the Lattic program run. Figuring out what the program does from a verbal description is going to be rather difficult. So, there may be a problem here. My university Information Service, dear souls that I am sure that they are, blocks the transmission of EXE files. This has fairly recently become a policy adopted by many universities. I don't have any idea of how to cope with this difficulty.
Maybe the above, plus the previous discussion with Bill Powers will provide something of a "concrete" basis for discussion.
Bill Williams
[From Bruce Nevin (06.03.2004 17:18 EDT)]
Rick Marken (2004.05.1030)--
[...] Bruce Nevin (06.03.2004 09:02 EDT) says:
> I have identified specific kinds of variables such that, if control
systems
> are controlling them, cultural phenomena will emerge. These are
>
> 1. Controlling aspects of behavioral outputs ("manner of controlling").
> This can be used as means of
> 2. Controlling social affiliation. (Crudely: who are my allies, who are
not.)
> And this can be used as means of
> 3. Controlling a perception of self in context of others.But this is all part of PCT already.
Yes, that is what I said. The "but" shows that you, too, are reading more
into it than I have said. You thought I was demanding a change to the PCT
model.
The difference is not in the PCT model, it is in PCT simulations
exemplifying the model. For cultural phenomena to emerge in a simulation in
the way that I have suggested, the agents must perceive behavioral outputs
as among the variables in the environment. Of course living organisms are
capable of perceiving the actions of effectors, their own and those of
other agents. No one has thought of building that capability into a
simulation. That which is perceived may be subject to control.
Each individual agent can then perceive its own behavioral output as an
instance of what it may also observe in the behavioral outputs of other
agents, and and can likewise perceive in the behavioral outputs of other
agents instances of what it would observe in its own behavioral output if
it were controlling certain perceptions with certain reference values. In
this way, it perceives that the other agent is controlling those
perceptions with those reference values.
Taken at face value, this perception of the intention of the other agent is
a delusion. After all, it hasn't performed the Test to identify the CV and
its reference value. The perception is confirmed when the agent is able to
control successfully in coordination with the other agents by controlling
its perceptions of the intentions of the other agents. These perceptions
(attribution of motivation) are really no more than projections of what it
would be intending if it were doing that.
These projections are made more reliable by conventionalizing ways of
behaving, sometimes actually as signals of intent, and by
institutionalizing the conventions in a public. An example of the first is
the ape baby nudging its mother for milk in a conventionalized way, but
this does not progress to institutionalization among more apes because it
is conventionalized anew by each mother-baby dyad (ape babies don't demand
milk from other mothers) and then the babies grow up.
This, too, is an emergent phenomenon, not adding anything new to the PCT
model, only adding to PCT simulations capacities for perception and control
that living organisms manifestly have.
/Bruce Nevin
At 10:31 AM 6/3/2004 -0700, Richard Marken wrote:
[From Bruce Nevin (06.03.2004 17:17 EDT)]
Bill Powers (2004.06.03.0747 MDT) sent 08:45 AM
Bill Powers (2004.06.03.0914 MDT) sent 09:25 AM
Bill Powers (2004.06.03.0931 MDT) sent 09:47 AM
We agree (I think) all four of us that individuals can in principle
acquire or establish control of “cultural” variables by the
trial-and-error process that Martin has described.
Three of us agree (I think) that this can also emerge through the process
that I have described, which involves controlling perceptions of one’s
own behavioral outputs in the course of controlling other variables,
controlling (recognizing) perceptions of others’ behavioral outputs, and
controlling perception of some sort of relation between them. You are
uncomfortable with this proposal.
You have assumed that the relation is identity (mirroring, mimicry), with
the objection that you can see your own face only in a mirror. The
relation between own outputs and other’s outputs is more
“categorial”. I use that word advisedly. I do not mean grouping
according to common perceptual features or according to resemblance to an
exemplar. “Cultural” perceptions seem to be a function of
making distinctions according to perceived differences. For example, in
language, there is no isolate category of voiced stops, say, containing
sounds like b, d, g. Rather, there is the contrast between voiced and
non-voiced, between stops and fricatives, etc.
Another example: that grimace on my colleague Derek’s face constitutes a
smile not by mimicry or direct likeness when it is compared to
(perceptions of) smiles on the faces of others, but by contrast with
(perceptions of) other facial expressions that I have perceived (could
now be perceiving but am not) on his face. His face was severely burned
when he was a child, but the same principle applies to unimpaired faces.
This is the basis of the difference between imitation and repetition. It
is also a key, I believe, to recognition under diverse conditions.
The way our language works, you cannot talk about a relationship without
prior mention of the relata, the elements that enter into the
relationship. Making a distinction, however, is a creative process in
which the relationship (the distinction) comes first and the relata (the
items that are distinguished one from the other) are derived from it. We
then turn around and delude ourselves that there are “things”
like p and b apart from the distinction that we (variably) make between
them. This leads to endless difficulties because the measurable sounds
and gestures are variable and their ranges of variation overlap. All that
is constant is the potential of having said the other instead, that is, a
perception of a relationship, a distinction.
When the behavioral outputs of two individuals are perceived as
equivalent, the set of distinctions is the same for both even though the
detailed means for making the distinctions may differ, and in that
context an x on one
side is perceived as equivalent to an
X on the
other.
An imitation is an attempt to produce the same perception in every
detail. A repetition is a production that is contrasted with
systematically distinct productions that are not made that could have
been (p vs. b, p vs. m, p vs. f, and so on). They are systematically
distinct within a field of conventional perceptual distinctions that have
been institutionalized in a public.
Of course I mean that a repetition is a behavioral output, perception of
which is perceived to be in contrast with perceptions of other behavioral
outputs … and so on. I do know that we are talking only about
perceptions, I just don’t want to get so garbled up with PCT-PC as to be
completely unintelligible.
BTW, there are lots of reasons for caution about using facial expression
too glibly for examples – the complexity of musculature and innervation,
the universality of emotional expressions cross culturally and even
across species. Your concern about the difficulty of seeing one’s own
face is uncertain ground for an argument. And anyway its a problem for
imitiation, not for repetition within a field of conventional
distinctions.
By either or both of those processes (reorganization is one, control of
perceptions of control actions is the other), individuals interacting in
a public each establish for themselves control of certain perceptual
variables with reference values such that the other individuals there
recognize (i.e. control) the perceptible effects as equivalent to their
own. When you say (or type) “you’re welcome” then Martin, and
Bill, and I each perceive that as the “same words” that I might
say (or type). Extranea like voice pitch or typeface are ignored because
they don’t make a difference between “you’re welcome” and any
other words. As you shake your head in disagreement, extranea such as
rotation about a horizontal axis from the bridge of your nose through the
back of your head are ignored by an observer, including yourself as
observer, because they don’t make a difference between shaking your head
in denial and nodding your head in agreement. You may recall that in some
other cultures head gestures for agreement and disagreement are
distinguished in a different way, and that for some of them rotation
about a horizontal axis from the bridge of the nose through the back of
the head is relevant.
Bill Powers (2004.06.03.0747 MDT) –
Bruce Nevin (06.03.2004 09:02 EDT)–
You are reading into it things that I am not
saying. I have identifiedspecific kinds of variables such that, if control systems are
controllingthem, cultural phenomena will emerge. These are
- Controlling aspects of behavioral outputs (“manner of
controlling”).This can be used as means of
- Controlling social affiliation. (Crudely: who are my allies, who are
not.)And this can be used as means of
- Controlling a perception of self in context of
others.This must be a very specialized definition of “self.” To me,
“myself”simply means me, this person, as opposed to someone else, that person.
Exactly so. “As opposed to someone else, that person.”
I don’t see that personal affiliations are
particularly relevant to who orwhat I am.
CSG is not relevant to who and what you are?
I don’t “affiliate” with myself, but
with others, and others aredifferent from me. That’s how I know they are “others.”.
I can’t make anything coherent out of these two sentences.
But I recognize
that some people do think affiliations are relevant in defining
themselves,which of course makes them relevant to those people.
None so blind.
When you say that controlling aspects of
behavioral outputs (meaning,controlling perceptions of them) can be used as a means of
controllingsocial affiliation, you’re automatically selecting just those persons
forwhom this is a goal.
Why do you jump from the potential of “can be used” to the
restriction to “selecting just those persons”?
And when you say that social affiliations
arecontrolled as a means of controlling a perception of self,
Why do you now change “can be used to control” into “are
controlled”?
you further
restrict the applicability of these ideas to individuals who define
themselves primarily in terms of how they relate to others.
Why do you add the word “primarily”?
So how big is
this final group? Are you saying that this specific hierarchy fits
everyone?
What specific hierarchy? As Rick pointed out, these are all capacities
within the hierarchy that is used for everything else. Nothing is being
added to the PCT model. I’m only supposing that a simulation of social
phenomena should include the degree of self awareness that social
organisms have. Or are you unable to perceive the sound and feeling of
yourself speaking and unable to perceive your gestures visually and
kinesthetically? As you correct typos in your reply, do you wonder why?
Does it annoy you that some others among us often do not? Why is
that?
Can you say “straw man”?
/Bruce
Nevin
At 08:45 AM 6/3/2004 -0600, Bill Powers wrote:
[From Bruce Nevin (06.03.2004 17:20 EDT)]
(Bill Williams 3 June 2004 1:00 PM CST) sent 01:38 PM --
If I understand correctly, each agent in your matrix program has these
perceptions:
A. Identity of four specific other agents (or three agents and a chalk mark).
B. Direction in which to perceive each of these four (front, back, left,
right).
C. Distance from each of these four, with a reference set by (D).
D. Relative size of distances in opposite directions (front vs. back, left
vs. right), with a reference of equality.
The environment is set up as a rectangle of chalk marks. The initial
position of the agents is "grouped in the middle" -- which could be
anything from a scramble to a compressed rectangle with everybody in the
right direction relationships and everybody except the outside members
controlling the relative distance perceptions with no error.
Given those perceptions and reference values, the reference values for
distance are a universal side effect of control. There is no room for
cultural variation.
What Martin is talking about, and what I am talking about, are ways in
which the agents could come to control these particular perceptions with
these particular initial-state reference values.
Martin (and Bill, sort of) are saying that it's all reorganization, and
that perceivable regularities are all post hoc. I'm saying that perceivable
regularities are a controllable, learnable social heritage.
If I were a person on a field with that task, I would attend to some other
perceptions, like dividing the width of the field by the number of people
on a row to estimate intervals and likely end position, or looking for
straight-row segments and controlling for being part of a straight row.
There would also be communication between people ("Well, we know we're
going to be on that side, c'mon." "Pull up a little, Joe. Don't worry,
they'll figure it out.") But of course this isn't a model of people on a
field. As I understand it, it's a study of the effect of disturbances on
interdependent control systems.
/Bruce Nevin
[Martin Taylor 2004.06.03.1658]
[From Bill Powers (2004.06.03.0931 MDT)]
Martin Taylor 2004.06.03.1049 --
I emphasise the "theoretical fact" that reorganization from infancy
onward, without the need to perceive our own output, will lead us to
develop actions
that effectively allow us to control our perceptions. In a social
context, that will happen optimally when our actions are sufficiently
like those of others in the same cultural group to influence them to
do what we want to have done.I agree, of course, within the loose construction of cause and effect. But
I don't think similarity is necessarily the key.
Where did the notion of "similarity" come into this? Perhaps from my
term "sufficiently like?" If so, please don't think that "similarity"
is a controlled perception, any more than "arc formation" is a
controlled perception in the Crowd program. Similarity occurs because
many of the people around you will react similarly to your saying or
doing X as they would to someone else saying or doing X in similar
conditions.
I don't suppose you remember my 1993 talk on this (it doesn't seem
nearly so long ago, to me:-). My point then wasn't so much to argue
for the perceptual reality of language as to argue that no matter
whether one started with a perfectly formally described language,
after a very few generations, it would have drifted into something
very hard to describe formally, because new learners (babies growing
up) would learn the language only in the sense that using it would
allow their perceptions to be controlled. And that would not mean
dotting every "t" and crossing every "i" [sic!]. Close enough is good
enough.
Bruce's daughter would be in much the same situation, when joining a
new peer group. Unless she shiftewd her way of talking, she would not
get from them what she would have needed to control perceptions
relating to freindship and feelings of belonging.
That's the area in which mimicry may well come into the picture, as a
quick way to get into the region where one's actions are good enough
to control to some degree the perceptions relating to the actions of
others. Sure, one's own voice sounds different on a recording to the
way it sounds live, but one can hear whether one is saying "barth",
"bahth", or "bath" (with a short-clipped "a"). One can tell whether
one is referring to a "lorry" or a "truck". These are things that can
be mimicked reasonably successfully. One's voice is more difficult.
Maggie Thatcher spent a lot of time in elocution lessons to learn to
speak with a lower voice. It's hard, and for most purposes, it
doesn't matter in regard to controlling your perceptions of how the
people around you respond to what you say.
When I want something I
don't say "You're welcome," but "please." The _other_ person says "You're
welcome." I think perception of the social variables is the key. We vary
our behavior in relation to other people to control social variables,
learning what works as we go.
I think you are being more "cognitive" than I want to be. I'll allow
that what you say can happen, but it's pretty sophisticated, and is
likely only to be used much by someone intellectual who _really_
wants to be taken to be a born member of the group.
.. My reservation is that we do not have nearly
as much reliable knowlege about how "people" behave in social situations as
one would guess from reading the literature.
No, but some men can guess pretty well how particular people, such as
their wives, are likely to react in certain situations. Less
particularly, although not everyone will move aside from blocking
your way if you say "Excuse me", most will, and when someone won't
you tend to ask yourself whether there is an environmental reason or
a personal reason why they won't. If you don't know the custom, you
might just try to barge through, and then you would quite probably
find the person seriously attempting to block you, rather than doing
so accidentally.
If you sat in a small group with one leg crossed over the other knee,
would you expect anybody else in the group to be insulted? I hear
that to show the sole of your foot is indeed an insult in some Arab
cultures. If you wanted something from an Arab in such a culture, it
might not be the best policy to sit so, and a child would learn that.
A Western visitor might have a hard time learning it if not told so
explicitly, but a very observant visitor might see than nobody else
in the group ever sat that way, and might unconsciously refrain.
Dyads in conversation often pose similarly, both with hands behind
the head, for example. It's quite non-conscious, until someone points
it out. I seem to remember some research that showed that people tend
to have friendlier feelings toward a conversational partner that does
that kind of mimicry than to one who sits differently. And that is a
kind of mimicry that can be done by perceiving one's own actions.
Whether it helps in controlling the perceptions one wants to control
through the actions of the partner is a quite different question. But
it might, and if it does, "to mimic" might well be an action used in
support of control of many different kinds of perception.
Enough ramble. All this stuff about mimicry is more Bruce's bag than
mine. My angle is that ordinary unconscious reorganization will, as a
side-effect, tend to bring one's language and behaviour into patterns
quite similar to the patterns used by the people with whom one most
often deals. And those patterns are what a sociologist sees as
cultural or linguistic norms, whether or not the actors know they are
following the norms.
Martin