Bowlby

[Bruce_Abbott]

[From Bruce Abbott (951012.1050 EST)]

Bill Powers (951012.0530 MDT) --

I find that the Bowlby book is available at Fort Lewis College; you
didn't tell me it was a 3-volume monster! I will get started on it
sooner than anticipated. There's a later book there, too.

Yes, but Bowlby's beautiful descriptions of control to which I referred are
found in the first volume, if memory serves. I read Bowlby first and later
encountered Powers. Instant synergy!

It's been a long time, though, since I last gazed upon those pages; perhaps
a second read would be in order now. Bowlby was a British psychiatrist
brought up on Freudian theory, which states that experiences very early in
life strongly influence the course of psychological development. However,
clinicians traditionally talked to adults and from these interviews
attempted to reconstruct those early experiences that had presumably led to
the development of certain aspects of the adult's personality. Bowlby
thought a better strategy would be to study children directly, and this
approach ultimately led him away from Freud and into the study of attachment
and loss that led to his publication of these volumes, now considered
classics in the field.

Regards,

Bruce

[William_T_Powers4]

[From Bill Powers (951025.1435 MDT)]

Bruce Abbott (951025.1045 EST) --

     Bill, I hope you have not given up reading Bowlby; what you had to
     say thus far could have been extracted from Chapter 1, which simply
     sets up Bowlby's approach without getting into the real nitty-
     gritty of it.

I was reflecting on why it is I find these materials so difficult to
read. The answer I seem to get is that I am not really very interested
in endless recitations of how people behave under various circumstances.
This is a terrible thing for me to admit, seeing how adamant I am about
experiments and demonstrations, but it probably accounts for some of our
differences.

A person like Bowlby is interested in control theory mainly as a
launching-point for a broad survey of human behavior in all its richness
and complexity, as it relates to the subjects of interest such as
attachment and loss. A person like me, on the other hand, considers one
example of behavior as good as another as a way of ferreting out
theoretical principles, with no particular example being of much
interest in itself. When you've seen one example of attachment, you've
seem 'em all. This, of course, is very unsatisfactory to a person whose
principal interest is in the phenomenon rather than the theory it
illuminates. But to a person whose primary interest is in understanding
the underlying organization of a system, the only reason for discussing
surface phenomena would be if they suggest something new about the
theory under construction.

At one level, I appreciate Bowlby's discussions of attachment, loss,
grief, mourning, and so forth simply as commentaries on life and as
insights into my own life. But at another level, I find them
excruciatingly repetitious and boring. To a theoretician like me, most
of the text is like an endless desert, with occasional oases which, on
closer inspection, turn out to be familiar territory and thus
unrewarding to visit again. I keep wanting to say to the author, "Wait,
stop here, you're skimming over something that could be important,
you're rushing on without pausing to reflect on how this thing works."
But of course authors can't hear such pleas from the reader; they press
on toward whatever point they are trying to make, leaving only their
camel-tracks on the spot where I would like to dig.

As a theoretician, I judge Bowlby not on the quantity of his data but on
the quality of his theory. Once you have the right theory, all the data
become uninterestingly predictable. Bowlby certainly went farther than
anyone else I have read toward a theory like PCT. He recognized the
process of control as central to all behavior; he saw how control
systems can come into conflict; he saw something of the hierarchical
structure of behavioral control systems.

But basically he didn't grasp how control really works, particularly the
way it is organized around perception rather than action. He took his
model from the TOTE unit. Here is how he summarizes on p. 18 of
_Attachment_:

     In place of psychical energy and its discharge, the central
     concepts are those of behvioural systems and their control, of
     information, negative feedback, and a behavioral form of
     homeostasis. The more complex forms of instinctive behavior are
     regarded as resulting from the execution of plans that, depending
     on the species, are more or less flexible. Execution of a plan, it
     is supposed, is initiated on receipt of certain information
     (derived by the sense organs either from external sources or from
     internal sources, or from a combination of the two), and guided,
     and ultimately terminated, by the continuous reception of further
     sets of information that have their origin in _the results of the
     action taken_ (and derived, in the same way, by the sense organs
     from external, internal, or combined sources).

His basic concept of control (the exact reference to which I can't find
at the moment) is that one set of sensory inputs starts the behavior,
and another set of sensory inputs, influenced in part by behavior,
terminates the behavior. The concept of the _closed causal loop_ eluded
him, as did the concept of controlling not behavior but perception. He,
like many others, was seduced by Miller, Galanter, and Pribram into
seeing control as a sequential process triggered on by one event and
triggered off by some consequence of the behavior. We, of course, can
see how close this idea comes to the essential organization of PCT, but
we can also see how fatally far from PCT it remains.

In _A Secure Base_ (I just found the reference) Bowlby says

     With supplies of physical energy available to them, these systems
     become active on receipt of certain sorts of signals and inactive
     on receipt of signals of other sorts. (p. 33)

So it's clear that he missed the basic point, the way a control system
reacts to deviations of its inputs from reference states by doing those
things required to restore them to the reference states. This gets
confused because (as in the passages you have just cited) he gets onto
control processes which are specifically sequential, practically the
only level of control that does more or less fit the TOTE model. So he
confuses control of sequences of perceptions with the production of
sequences of acts -- which, of course, do occur when perceptions are
created in sequence. However, when disturbances occur, different actions
are generally required to maintain the same sequence of perceptions, so
if you fail to notice that it is perceptions that are under control, you
may be misled by seeing a change in the sequence of actions. For
example, if a bee fails to get the right perception in sequence, it may
simply switch to another flower instead of continuing with the next
behavior that would get it to the nectar on the same flower. But it's
still trying to control the same sequence of _perceptions_.

Often Bowlby seems to say things that come much closer to PCT, but that
is an illusion, because while we can see their significance, he does not
seem to do so -- he doesn't follow up the obvious deductions. He sees
that the behavior of attachment has the effect of creating proximity to
the object, but he doesn't see the obvious follow-up that there is a
reference level for proximity, and the behavior we see is simply the
result of the error. He didn't see proximity as simultaneously acting in
the roles of cause and effect. When proximity already exists, he doesn't
see the same mechanism working, simply because there is no error and
thus no action required to correct it. Of course _we_ can see that, but
Bowlby doesn't. He doesn't see the _satisfied_ control system as being
the same as the _unsatisfied_ control system. His orientation is toward
seeing behavioral _events_ following from environmental _events_. It is
a deep-seated S-R orientation, very hard to get rid of. That's the
appeal of the TOTE unit.

When a child wants the mother to be near, the child is _always_ engaged
in getting the mother to be near, _even when the mother is in fact
near_. The presence of this control system isn't obvious when its error
signal is very small, but all that is required is a disturbance to bring
it to life again; in truth, it was always active. So when you know that
this reference level exists, you also know how many different kinds of
detailed situations will affect behavior -- all those situations that
have the effect of creating a nearness error, by whatever means. The
disturbances can take many forms, such as a stranger intervening or
taking the child away, or the mother becoming occupied with some task
having nothing to do with the child, or the child being distracted and
losing track of the mother. These disturbances can look like very
different environmental events, which makes the behavioral responses
look different, too, but when you understand that nearness is under
control at all times, you can see that all these different-looking
responses are just the same control system doing the same thing by
different means.

This is what I mean by saying that when you've seen one example of
attachment you've seem 'em all. The variety of environmental situations
and the variety of behavioral "responses" to them is far greater than
the variety of controlled variables and systems for controlling them.
We're just seeing the same control process with different disturbances
acting. Of course I'm saying that hypothetically, as if we have done the
digging at the right spots and have actually found the variables that
the child is controlling. But that is why I keep wanting the author to
stop and dig, instead of going on to enumerate more disturbances and
more behavioral responses to them. You don't need to enumerate every
crosswind, every bump and tilt in the road, once you understand that the
driver is keeping the car in its lane by turning the steering wheel _as
required_.

I think Bowlby came very close to inventing PCT. If he had been less of
a clinician and more of a theoretician, he might have put it all
together. Perhaps you're on the right track in using Bowlby as a bridge
to conventional psychology and ethology; PCT is certainly much less of a
break with his work than with most other work.

···

-----------------------------------------------------------------------
Best,

Bill P.

[Bruce_Abbott]

[From Bruce Abbott (951025.2005 EST)]

Bill Powers (951025.1435 MDT)

At one level, I appreciate Bowlby's discussions of attachment, loss,
grief, mourning, and so forth simply as commentaries on life and as
insights into my own life. But at another level, I find them
excruciatingly repetitious and boring. To a theoretician like me, most
of the text is like an endless desert, with occasional oases which, on
closer inspection, turn out to be familiar territory and thus
unrewarding to visit again. I keep wanting to say to the author, "Wait,
stop here, you're skimming over something that could be important,
you're rushing on without pausing to reflect on how this thing works."
But of course authors can't hear such pleas from the reader; they press
on toward whatever point they are trying to make, leaving only their
camel-tracks on the spot where I would like to dig.

Although at one level I can appreciate your feeling, on another I can't help
but think that you've skimmed over much of what Bowlby had to say in the
middle chapters and that this is why you are offering vague generalities in
the above paragraph--you could have said these things without reading
anything at all. In particular you seemed to have missed some significant
passages that might change your view of what Bowlby did and did not
understand about control systems. You say:

His basic concept of control (the exact reference to which I can't find
at the moment) is that one set of sensory inputs starts the behavior,
and another set of sensory inputs, influenced in part by behavior,
terminates the behavior. The concept of the _closed causal loop_ eluded
him, as did the concept of controlling not behavior but perception. He,
like many others, was seduced by Miller, Galanter, and Pribram into
seeing control as a sequential process triggered on by one event and
triggered off by some consequence of the behavior. We, of course, can
see how close this idea comes to the essential organization of PCT, but
we can also see how fatally far from PCT it remains.

Bowlby had indeed read Miller et al. and understood the theoretical
importance of the hierarchically-organized plan. But he did not necessarily
see the individual component systems of a plan as TOTEs. For example,
Bowlby has this to say:

    Past discussions of instinctive behaviour have tended to concentrate on
    sequences that have a dramatic and time-limited outcome, and to neglect
    behaviour the outcome of which is an ongoing relationship, such as
    maintenance of a specified distance over a comparatively long period.
    There can be no doubt, however, that the latter sort of behaviour is of
    great frequency and vital importance in the life of most animals.
    Examples are brooding behaviour, in which proximity to eggs and nest is
    maintained over weeks, territorial behavior, in which location within a
    certain bit of the environment is maintained over months and sometimes
    over years, and alarm behaviour, in which a certain distance between an
    animal and a predator threatening it is maintained over minutes or hours.

    A main reason for past neglect of this sort of behaviour is that it cannot
    readily be understood in terms of such concepts as 'drive' or 'energy
    discharge.' When the present set of concepts is applied, on the other
    hand, possible solutions are not too difficult to see.
        Bowlby (1969, p. 72)

On the next page, writing on the orientatin of behavior, he adds this:

    In the case of the peregrine's stoop and all other goal-corrected behavior
    in which the set-goal includes a specified part of the environment to or
    away from which the animal moves, orientation results from a continuous
    comparison of performance with set-goal. In those cases the system
    responsible is working on a 'homing' principle, as in the case of the
    homing missle.

    In other closely related cases the animal itself does not move, but
    either its whole body or some part of it, for example eyes, is turned
    towards an object in the environment. In the cases the system
    responsible may be working on a 'tracking' principle, as in the case of
    a radar receiver set to lock onto and track an aircraft. An example of
    tracking followed by an aimed movement is the way in which a toad
    catches a fly by shooting out its tongue towards the fly. Here
    orientation is achieved by the toad's turning its head towards the fly
    in a tracking movement. The tongue itself, on the other hand, does no
    more than extrude straight from the frog's mouth and is oriented in no
    other way. Systems responsible for this form of orientation are working
    on the same principle as a man-aimed gun. In each case the aim is a
    result of movement controlled by a goal-corrected tracking system, and
    the subsequent movements of tongue and bullet result from the action
    of a simple fixed-action type of system.

None of these cases resembles the test-operate-test-exit program outlined by
Miller et al.; all are examples of continuous error-corrected systems.

Bowlby does offer a section at the start of Chapter 6 headed "Activation and
termination of behavioural systems," which is followed by the subheading
"Starting and Stopping: Classes of Causal Factor." Here he considers how
behavioral systems are brought into play--in effect why, with the limited
number effectors available, one system's set-goals are pursued at a given
moment as opposed to another's, when both cannot be pursued at once. My own
example would be the rat which is first observed to eat and then to drink,
then to eat again, then to run in the running wheel. The rat cannot perform
all these activities simultaneously; what determines which system will be
"active" in the sense of having setting the reference levels for the
lower-level systems? The fact that Bowlby concerns himself here with
starting and stopping is not an indication that he views these systems as TOTEs.

In _A Secure Base_ (I just found the reference) Bowlby says

    With supplies of physical energy available to them, these systems
    become active on receipt of certain sorts of signals and inactive
    on receipt of signals of other sorts. (p. 33)

So it's clear that he missed the basic point, the way a control system
reacts to deviations of its inputs from reference states by doing those
things required to restore them to the reference states. This gets
confused because (as in the passages you have just cited) he gets onto
control processes which are specifically sequential, practically the
only level of control that does more or less fit the TOTE model. So he
confuses control of sequences of perceptions with the production of
sequences of acts -- which, of course, do occur when perceptions are
created in sequence. However, when disturbances occur, different actions
are generally required to maintain the same sequence of perceptions, so
if you fail to notice that it is perceptions that are under control, you
may be misled by seeing a change in the sequence of actions. For
example, if a bee fails to get the right perception in sequence, it may
simply switch to another flower instead of continuing with the next
behavior that would get it to the nectar on the same flower. But it's
still trying to control the same sequence of _perceptions_.

As I think I've demonstrated, Bowlby was more sophisticated in his
understanding of control systems than this passage suggests. When he deals
with sequential control processes, it is because he is attempting to account
for sequential patterns of behavior, as in behavioral chains.

I am not claiming that Bowlby had an engineer's understanding of control
system dynamics, but I do believe, based on my reading of him, that he
understood how such systems might be organized to produce the phenomena of
attachment with which he was concerned. In developing his view, he
considers a number of problems and observations that do not seem to me to be
even raised, much less answered, by PCT.

Regards,

Bruce

[Bruce_Abbott]

[From Bruce Abbott (951027.1230 EST)]

Bill Powers (951026.0530 MDT)

Before we discuss your interpretation of Bowlby's views on the relationship
of biological function to controlled perceptions and their references, we
should first introduce what Bowlby himself had to say. In _Volume 1:
Attachment_, there is an entire chapter entitled "Function of Instinctive
Behavior, from which the following is extracted:

    It has been emphasised in this chapter that the existence in an animal
    of a particular species of any one environmentally stable behavioural
    system comes about because the activity of that system commonly has a
    consequence that is of survival value to the species. Systems
    responsible for eating behaviour commonly have food-intake as a
    consequence. Systems responsible for mating behaviour commonly have
    reproduction as a consequence. Since the activities of these systems
    so commonly and so plainly fulfil a biological need, why then not term
    them 'need systems'?

    There are at least three good reasons for not doing so. First, in each
    instance the activity of the behavioural system in question may have
    consequences of quite other kinds. In a particular individual a system
    that is obviously connected with food-intake may have as its main
    consequence the sucking of a thumb or a pipe. In another individual
    a system obviously connected with mating may have as its main consequence
    sexual activities directed towards a fetish or an individual of the same
    sex. In such instances the activity of the system has no survival value
    whatever. To call such a system a need system is therefore confusing;
    and the confusion is only made worse if, to meet the difficulty, new
    needs, e.g., to suck, are postulated. Secondly, as already remarked,
    there are a number of species-characteristic behavioural systems the
    biological functions of which are still unclear. This fact is obscured
    when every behavioural system is termed a need system, because the term
    'need' tends to imply that the usefulness of the system is self-evident.
    Finally, the term 'need system' can readily lead to an assumption that
    the need plays some sort of causal role in activating the system, the
    fallacy of teleology.

    A legitimate usage of the term 'need' is to restrict it to refer to the
    requirements of species survival. If the species is to survive, an
    animal can be said to need food, warmth, a nesting-site, a mate, and so
    on. Obviously none of these needs is a behavioural system, nor does
    any of them cause the activation of a behavioural system. On the other
    hand, many behavioural systems have the function of meeting one or another
    of these needs; and it is because, if that species is to survive, those
    functions have to be fulfilled that those particular behavioural systems
    have been evolved. Needs, therefore, are not the causes of instinctive
    behaviour. What they do is determine the functions that behavioural
    systems have to serve. Thus they constitute the selection pressures
    under which behavioural systems evolve.
        Bowlby (1969), Pp. 137-138.

    A couple of paragraphs later, Bowlby has this to say about the uses of
the terms "purpose" and "aim":

    A difficulty about both 'purpose' and 'aim' is that each tends to carry
    overtones of teleological causation. A more serious difficulty about
    them, moreover, is that each is habitually used in ways that fail to
    distinguish between a system's predictable outcome and its function--a
    fatal confusion. For this reason neither is used in this work.
        Bowlby (1969), Pp. 138-139.

For comparison, here are your remarks about Bowlby's views on this matter:

One important thesis is in the earlier books and the later one as well:

    The distinction I have drawn between the function served by a
    certain form of behavior and our knowledge of, and our striving to
    reach, the conditions that will terminate that behavior is one of
    the criteria that distinguish the biological realm from the
    psychological (A Secure Base, p. 64).

Note that "the conditions that will terminate the behavior" is the same
as the PCT definition of a reference level, although there is no sign
that Bowlby took this idea any further.

Here he distingushes between what I would call learned control systems
and intrinsic ones. However, he doesn't exactly make the distinction
that way. He is saying that when the behavioral systems seek a goal, the
effects of achieving it ("terminating the behavior") are to be judged in
terms of the biological function served by attaining that goal, not by
the goal itself. This is similar to my concept of a reorganizing system,
in that the reason for acquiring any behavioral control system is
ultimately to maintain certain internal critical variables near their
inherited reference levels. But it differs from my concept in a
fundamental way.

Bowlby does not cast the deeper reason in terms of more control systems;
he is saying that the reason for existence of a behavioral goal is its
biological function AS THE BIOLOGIST SEES IT. For example, we have the
goal of eating food containing vitamin D because without vitamin D we
will develop a vitamin deficiency disease and become less able to
reproduce ourselves. This puts the "because" into the realm of objective
facts about the organism, as observed from the viewpoint of an
independent observer, and throws the burden onto evolution, the catch-
all explanation

In PCT we don't use such "becauses;" what the external observer knows
about the organism has no influence on how it works. It may be true that
if the organism doesn't eat, it will die, but the organism doesn't know
that, and this fact about not eating is not what makes the organism eat.
What makes the organism eat is a reference signal for a state of
nutrition, being compared against a perception of the actual state of
nutrition, or something like that -- not an abstract fact. Organisms
lacking this intrinsic control system will not eat, and sure enough,
they will die. But if they eat, and live, it is not because they will
die if they don't eat, but because they contain this control system.

The way biologists speak of functions bypasses the means by which these
functions are accomplished. This creates a very peculiar mode of
explanation, in which the consequences of a certain kind of behavior are
offered as a direct explanation for why that behavior occurs. The
biologist says that the goose pulls the egg back into its nest because
this improves the chances of survival of the species. That's the
functional explanation of this behavior. However, although it is true
that if geese do not pull stray eggs back into the nest the survival of
the species will be adversely affected to some degree, it is not true
that they retrieve the eggs _because_ doing so will promote survival.
They retrieve the eggs because of the operation of a specific control
process inside of them. As a consequence of the existence of this
control system, the species which commonly contains it will have a
better chance of survival; the effect on survival is a consequence of
the operation of this control system, not a cause.

When future consequences of behavior are cited as causes of the behavior
that leads to them, one is really talking about a control process. But
one has to be careful that in doing so, one is not assigning the
consequence the role of a reference signal. The consequence follows from
behavior, but the reference signal does not. The reference signal is a
prerequisite for reaching the goal-state.

Furthermore, as we have been saying recently on the net, not every
consequence of behavior that we can notice is necessarily a controlled
variable. Geese that control the positions of their eggs certainly
control the positions of their eggs, and that is a controlled
consequence of their behavior. But the same behavior has an uncontrolled
side-effect, which is to improve the chances that there will be a
hatchling from the egg, and as another consequence that the population
of geese will be higher than it would be if the position of the egg was
not controlled. Those side-effects are not under control by the goose at
any level of its organization. They are not _reasons for_ its control
behavior, but _consequences of_ its control behavior.

So biological functional explanations are really not explanations at
all; they are simply observations of consequences. The wolf marks its
territory, and other wolves stay away (sometimes). The stickleback darts
at an object with a red mark on it, and the object moves away. The chick
begins to follow a large warm object, and follows that object to food or
water or safety. But as far as research can prove to date, the wolf does
not mark its territory IN ORDER TO drive other wolves away; the
stickleback does not dart at the red mark IN ORDER TO drive a rival
away; the chick does not follow the mother IN ORDER TO get food, water,
and safety. What follows "IN ORDER TO" is simply a consequence of the
observed behavior -- until we can establish that these consequences are
_intended by the organism_. With the Test, we can do that, or else prove
that the consequence is merely a side-effect of controlling something
else.

The mistake that the functional explanation makes is to assume that ALL
important consequences of behavior are intended; that they all work like
reference signals. This is clearly not true, and can't be assumed _a
priori_. It is certainly true that maintaining a body temperature of 37
C makes our biochemical processes run at a favorable rate, but it is not
true that we maintain this temperature _because_ it does this. As far as
I know, we have no sensors that can detect the rate of a biochemical
reaction, and no reference levels for such rates, either.

So I disagree, on PCT grounds, with Bowlby's distinction. I say it's
control all the way down, and that functional analysis is irrelevant in
explanations of behavior.

Bill, it seems to me that you have just repeated Bowlby's argument. I
haven't read _A Secure Base_, perhaps by then he had changed his mind (or
perhaps you have misunderstood his meaning), but he clearly was on the right
track in _Attachment_. It's really a pity that you and he never got together.

Regards,

Bruce