>What is the relationship between dialogue and activity, anyway? I got the
>point that it involves cutting and pasting. But is there more to it?
>mike
To apologize in advance (I know I shouldn't but I will), this thing I am
writing is going to be long - its about 5,000 words and its length is way
out of line for a discussion list. It's not a post at all, it's an essay.
I promise that I don't intend to normally write such lengthy pieces -
honest! - so please don't get too wrong an impression. I am new to this
list and can only get away with doing something like this once. And I know
some won't have time to read this or won't care. But I would really like
to indulge in some analysis of some of the issues Wells brings up, since I
have been studying his paper, and I am really enjoying working with these
ideas. If someone does read this and have a comment or two, I would be
most appreciative.
The short answer to Mike's question above that I am proposing in my
commentary on Wells' paper below is that dialogue and physical activity are
inherent components of all human activity, and can be modeled by a unified
plan-do activity process cycle.
Commentary on the Wells paper by Steve Gabosch Oct 28, 2002
Summary
This commentary agrees with Wells' proposal to see both sign and tool
mediated activity as similar but different, but it suggests advancing this
concept to the next level. Rather than seeing them as two different kinds
of human activity, it proposes a model (to give it a name I'll call it the
plan-do activity process model) where both sign-mediated and tool-mediated
activity are inherent parts of any human activity process. It suggests 3
"planning" steps (get info, think and talk) and 3 "doing" steps (decide,
act and reset) are components of human activity in general. In this model,
both material things and abstract objects (meanings) are the objects being
acted upon, incorporating both tool- and sign-mediated activity into the
model . An advantage of this model is that it solves one of the key
problems Wells is addressing - how to incorporate more than one subject in
an activity process. According to this model, if only one person is
involved, their thinking process becomes a special form of discourse -
thinking to oneself. This model proposes two measurable units of analysis
to help analyze an activity cycle - the decision, and the outcome of the
cycle. These two units of analysis, the decision and the outcome of the
cycle, can be considered "deliverables" and can be measured in terms of
their concrete manifestations (for decisions, words and/or physical
movements; for outcomes, physical changes in objects, including the
subject's body). An advantage of this model - if it works the way I think
it can - is that it can be applied to a large series of activities, such as
a project, a sequence of tasks, or just a single step. This commentary
spends some time analyzing the Janet-Linda dialogue using this model. It
ends up with a description and depiction of a diagram of this plan-do
activity process model I have in mind, based on the diagrams published by
Wells.
Commentary
In his paper "The Role of Dialogue in Activity Theory" (Mind, Culture and
Activity, 9(1), 43-66, 2002), Gordon Wells is wrestling, while making many
good points, with several traditional problems in philosophy and scientific
methodology: the complex relationships between subjects and objects, signs
and meanings, dialogues and physical activities, semiotic and material
activities, and so forth. .
For example, this passage in Wells' paper sums up his thoughts on the
relationship between material and semiotic actions:
"Thus, material and semiotic actions should not be thought of as mutually
exclusive alternative forms of joint activity. Frequently, they occur
simultaneously or alternate as phases in the same activity; in either case
they are in important ways complementary. As well as distinguishing the
different modes in which tools and signs mediate activity, therefore, it is
equally important to try to understand what might be called their
"intertextuality" (Lemke,1995)." (page 50).
I am not sure what "intertextuality" means, but I like the direction Wells
is trying to go. However, Wells seems ambiguous on some of the relations
between material and semiotic actions, which are highly dialectical and
contradictory. My thinking is to take Wells' proposal even farther than he
does. I suggest that the material and the semiotic are not just
"important," they are completely and irrevocably necessary and indeed
"intertwined," as he refers to them. I suggest that both material and
semiotic actions are mutually and collectively essential in any activity
process, and that it is necessary to track both aspects to scientifically
understand an activity. Wells grasps this but seems shy about just saying it.
My thinking sees both an abstract and a concrete aspect to all the key
entities in human activity - the object being acted on, the tools, the
artifacts, the words spoken or written, etc. Just as material objects can
be considered concrete and semiotic objects can be considered abstract, I
would argue that there is an abstract aspect to any material object (its
meaning) and a concrete aspect to any expressed semiotic object (the
utterance, sign on a page, etc.). This means that all material objects
have a semiotic aspect and all semiotic objects have a material aspect.
The one exception to this rule might be pure thinking, where a person is
keeping their thoughts to themselves and not expressing them externally.
This could be considered a special case of discourse, where inner speech,
invisible to others, replaces external dialogue, and abstract objects are
only processed in the mind of the thinker. To my knowledge, this way of
thinking is classic Vygotskyism, and is basic to Activity Theory.
Wells seems to be uncertain of this way of looking at the relationship of
subject(s) and the objects in the activity process. On page 48 he quotes
from e-mail from Arne Raeithel to introduce the notion that "the
distinguishing mark of sign mediated action in comparison with tool
mediated action (narrow sense) is precisely that the object of the activity
is the subject itself. Subject acts on Subject via mediational means."
Wells refers to Figures 3 and 4 (by Raeithal) in his paper as depicting
"the essential difference between object-oriented action and
subject-oriented action."
This approach seems to confuse the relationship of subject and object, sign
and tool, etc., in human activity. Wells seems to be imagining some kind
of dualism between the semiotic and material that understands that they are
different but is unclear on how they are also unified. And he seems to
lose sight of the notion of subject and object. Wells wants to show where
sign-mediation comes into the picture, but loses ground when he implies a
model of activity that while it affirms that tool-mediated activity acts on
objects, at the same time, however, it identifies sign-mediated activity as
acting on subjects. (In a sense, the solution to this problem is simple
and almost semantic - sign-mediated activity could be said to work on the
abstract objects (thoughts, etc.) in another person. Perhaps this is all
Wells means - but he doesn't exactly say it.)
Following what I understand so far from Vygotsky and the advocates of
Activity Theory - I suggest that a model of the activity process that
retains the traditional Vygotskyist concept of subject and object, but at
the same time unifies both kinds of activity - material and semiotic -
while also clearly distinguishing between them as the opposites they are.
In this model, both sawing a piece of wood in half and having a
conversation with a neighbor about the weather should be equally
understandable as both simultaneously containing material and semiotic
activities within a unified activity process.
According to this model, in any instance of human activity, there is a
planning sequence followed by a doing sequence. Intuitively, I would rough
out six steps altogether: getting information, thinking, talking, making
or agreeing to a decision, acting, and resetting. The first three, getting
info, thinking and talking - the planning sequence - can loop around for
some time. Finally, beginning the doing sequence, a decision or agreement
is made, and then physical activity on the object in question is set in
motion. After performing the activity, there is usually some kind of
physical disengagement, and then the cycle can repeat under the new
circumstances.
There is nothing unusual about these six steps - nor should they be, if one
is to claim they are a general schematic description of human activity.
For the purposes of this discussion, they are only a rough first try at
defining them. Intuitively, they seem to feel right, but that is not a
scientific claim. Lists of steps in an activity like this are common in
industrial engineering process analysis, but that is also not the same as
empirically-based scientific theory.
Interestingly, Arne Raeithel touches on this idea of an activity cycle
sequence in one of the emails from that discussion Wells refers to that can
be found at ftp://weber.ucsd.edu/pub/lchc/chapters/april95prac (found
thanks to Google).
Raeithel says:
"In German Work Psychology the concept of cyclic completeness of an
activity has been proposed by Winfried Hacker. He has also empirical
data that cyclic completeness is healthy in a basic sense, and also
in more refined senses, like mental sanity, or non-breaking families.
"I am sure that some similar ideal phase description has been worked
out in many schemes for analysis. I will give my presently preferred
list:
* recurrence of demand, and initiation of activity
* orientation and preparation
* willing
* realising
* checking, delivering, and fade-out." (end of Raeithel)
I don't know what "recurrence of demand" means, but this list Raeithel
preferred in 1995 closely matches the sequence in this plan-do activity
model I am suggesting. The delivery idea in this list also figures in as
an important feature of the model I have in mind - the idea of deliverables
described below. None of this similarity should be any surprise - we are
talking about a simple description of something completely familiar to all
human beings - human activity. The real trick to getting this right is
sorting out the material and semiotic aspects of each part of the process,
and this is the place thinking people are likely to have important
differences of opinion over terminology and content. From my understanding
of Vygotsky, I believe this plan-do activity process model is consistent
with his methodology - but many don't agree with Vygotsky's methodology.
Of course, these simplified schemes are infrequently close to what human
activities often actually look like, although these stages (or something
like them) are generally there - maybe always there. Humans often make
their activity sequences look much more complicated than this simplified
scheme because they may scramble steps and act before they think, or talk
before they have good information, they may misunderstand things and make
execution errors, they may interrupt a cycle and start others before
finishing the first, they are constantly employing trial and error
techniques, they not infrequently have conflicts with other subjects, the
rules and conventions, or the community and social structure, and in
general wind up doing a lot of different activities and re-doing a lot of
different things seemingly all at once. This, of course, can be hard to
sort out if one is not familiar with the particular kind of activity they
are observing or participating in. This outward appearance of chaos
surrounding what is really a profoundly logical, dialectical process is one
of the things that makes humans so fascinating - at least to other human
beings.
Another feature of the complexity of human activity important to this model
is the human activity tendency to nest tasks within tasks - doing several
things at once, and completing them in a different order than they were
started. Tasks within tasks may in turn be part of larger projects. This
can be true of both concrete tasks (I am going to wash a dish, but first I
have to find the sponge) and abstract tasks (we need to select a single
project from many proposed projects, but first we have to discuss what
resources we have for each proposal). When a person or a group loses track
of nested tasks, confusion usually sets in.
According to this model, any project, such as making a model land yacht out
of classroom junk materials, consists of possibly hundreds of individual
activity process cycles. The object being acted on in any given cycle may
be a tool (cleaning it to use it), an artifact (finding it), the central
project object (assembling it), a logbook (writing in it), or a proposal
(discussing it and deciding whether to adopt it).
Also according to this model, two deliverables - concrete, measurable
behaviors and outcomes - are produced in each activity cycle. One is the
physical manifestation of the completion of the decision stage. This can
be a simple as getting up from a table to go do something you have just
decided to do, or as complex as signing a contract (or perhaps announcing a
vote total). The second deliverable is the concrete outcome of the
activities. This can be as simple as sawing through another quarter inch
of wood or constructing a model of a land yacht (or perhaps building a
skyscraper). As Wells points out, in principle, once a model incorporates
two subjects, many more can be added. I am not comfortable leaping from
what seems to start out as psychology to what seems to be sociology (votes,
skyscrapers) but once two subjects are let into the model, the barn door is
open, and as many as can do an activity together need to be accounted for.
(Who ever said this was easy?)
Let's turn our attention to the rich empirical data Wells provides. I
suggest that to understand the overall direction of particular activities,
such as the project Janet and Linda are described in, one needs to
understand the ultimate project deliverables that the subjects perceive
they should be working on. In this case, the project deliverables for
Janet and Linda were the logbooks and a suitable machine of some kind made
of classroom junk materials. The project deliverables can then be broken
down into the specific tasks necessary, each of which will have their own
deliverables. (Note that if the ultimate material deliverables were the
logbooks and the model, the ultimate semiotic deliverables were the
decisions to turn the logbooks in and present the model as finished).
The first task of the classroom project involving Janet and Linda was to
write in the logbooks and keep writing in them, the second was to choose
the machine they were going to build, and the third was to design it. Just
as all other objects and activities have their concrete and abstract
aspects, as argued above, deliverables also have both aspects. The
deliverables in the first task are the written logbook words. But these
concrete letters and words also have to be meaningful and appropriate
according to the teacher's criteria to get a good grade. In the second
task, the deliverable is a decision on what to make. But this abstract
object, a decision, also has its concrete aspect - it needs a picture, a
label ("boat on wheels"), and so forth to go with it so it can be explained
and talked about. The third task, creating a design, only gets started in
the data in Table 1. And this abstract object, the design, also has a
concrete aspect - the availability of the materials and tools and artifacts
that are called out in the design (such as a newspaper to supply wind power
and certain cloth to make a sail) - each abstract design proposal Janet and
Linda came up with had to be evaluated against the concrete constraints of
the allowable materials (junk found in the classroom).
Table 1 divides about 15 minutes worth of events into 71 Ref sections. One
of the problems here is that Wells has written up all this beautiful data,
but provides no explicit way to clearly distinguish what is significant
from what is not, except by his general commentary. This scheme or method
I am presenting - sorting the events in terms of plan-do activity cycles,
nested tasks, and deliverables - has the advantage of being able to sort
this data out according to a method. It can also break a project down into
middle-sized or unit-sized pieces. It can - or so I am claiming - look at
the project as a whole, look at any of the major sequences of tasks, look
at individual steps, or at any of the nested tasks using these units and
groups of units of analysis.
Wells divides these 71 Ref sections into 14 sequences, which pretty much
follow this plan-do activity analysis scheme I am presenting, but he does
not focus on why he groups these sequences that way, nor does he point
these groupings out with any emphasis. These groupings are fairly easy to
do intuitively, since everyone is familiar with the stages of human
activity, but to do it scientifically, one needs a conscious method (so
someone can come up with an even better method). In particular, we need a
way to sort out all the nested activities that are usually associated with
any task or project.
First, let's group the Ref sections in terms of the major deliverable of
the dialogue in Table 1, the creation of a decision of what the project
should be:
Ref 01 - 02. Start of discussion on what to make. Linda initiates the
discussion: "What are we planning on making?" Janet responds affirmatively
by reading her writing on the question to Linda. "I put \this ...".
Ref 03 - 60. Planning activities commence. These activities are comprised
of several nested tasks. They read out loud from and write in log books,
look at the illustrated book, make proposals, discuss some of the problems
and possibilities of these proposals, and interact with C to secure the
illustrated book when they believe they need it.
Ref 61 - 65. In this series of Ref sections, a definite decision is made
to do the boat. Linda initiates a conversation that goes, essentially,
"Have we decided to make that boat?", to which Janet responds, "I think we
should \make that," where Linda says "OK" and then Janet says "OK." The OK
from each of the girls signifies in concrete utterances their definite
mutual agreement. The "reset" stage is almost instantaneous and they
immediately begin their next task. (In fact, they seem to be ready to go
to the next task, design, even before they have fully completed saying "OK"!)
Ref 66 - 71(end). More discussion of boat and writing. Now, the focus has
changed significantly. The focus is no longer *whether* and *can* they do
the land yacht model - the focus now is *how*.
This overall decision-making task appears to follow the plan-do cycle.
First, Linda takes the initiative and defines the task - let's decide on a
project. They spend some time discussing and researching proposals, then
arrive at a decision, which is symbolized in an OK from each. Then they
move on to the next task.
We can also use this same activity process analysis to breakdown the stages
of a subtask. Let's take the series where Janet proposed the Egyptian
roller machine, but eventually the girls move on.
Ref 22 - 23. Log roller proposal. Janet makes a proposal based on a
picture "Linda!. This is neat ...", and Linda uses gestures (looks at book)
to show interest.
Ref. 24 - 30. J and L discuss Egyptian roller proposal.
Ref 31 - 33. Mutual decision to look further. Janet appears to be calling
attention to another picture on a facing page and Linda says "let's look
through some \more," and Janet responds "Let's see -" while turning page.
In this series, Janet makes a proposal, which gets discussed, but the girls
do not make a decision; they do not decide to go with the Egyptian roller.
Instead, Linda proposes to continue their general search and get more
information from the book. Janet completes this activity cycle begun in
Ref 22 by turning the page, where they both reset and begin looking at new
pictures. The decision deliverable was achieved by Linda's suggestion to
keep looking and Janet's internal decision to say "Let's see - " and turn
the page. The activity deliverable was achieved by Janet turning to a new
page and the two of them redirecting their attention. In a sense, the
outcome of this activity cycle was to abort the activity and move on - a
rather frequent event in human activities, to be sure.
In the next Ref section, begun in Ref 36 "Look at that sailboat," Linda
calls attention to the sailboat, which begins a new subtask - consider the
idea, which Linda motivated effectively with the selling point "wind power"
- and then the subtask is completed in R 64 - 65 when they both say OK.
The completion of this subtask then also becomes the completion of the
general task - choose a project.
Returning to Wells, his loose ends regarding analyzing Janet and Linda's
dialogue - providing only his intuitive comments, all interesting, but no
method of analysis - may be solved with this activity process sequence I
have described. By looking for six steps and two deliverables, we can
break a project down into smaller tasks sequences - even down to the most
specific steps if we want to. This seems to work with the data from the
Janet and Linda project.
Another loose end is the unsatisfying diagram in Fig. 5, which has little
or no movement represented. If a diagram of these 6 process stages is
placed inside Wells' diagram in Figure 5, a more dynamic representation of
the activity process can be depicted.
Borrowing most of the ideas in Wells' diagram, we can add some helpful
features. Here is a verbal description of this attempt. At the bottom of
the triangle are the major social forces and conditions that mediate the
context and the constraints for the subject(s) efforts (rules and
conventions, community, division of labor). Next is the subject or group
of subjects. Next is the current task or project, at the level of
specificity or generality desired by the analyst. Inside or alongside the
task, the 6-stage activity process is represented - get info, think, talk,
decide, do, and reset. At the end of each activity there is an outcome,
which changes conditions, calling for a new task, or repeating the one just
accomplished. Each time, a new pair of deliverables is planned and created
- there is a new decision, and a new outcome in each cycle. The activity
process cycles again and again as tasks and objects being acted upon keep
changing, new conditions keep appearing, and new decisions need to be made.
Arrows can help indicate this repetitive cycle. Surrounding the activity
cycle in this suggested diagram are labels for the key entities that
mediate the activity itself - the tools, artifacts, the central object in
process, etc. Above the six-step activity process is the outcome (which
might be considered a seventh stage because observing this outcome may be
the content of the new information starting the next cycle).
**************************************************
project outcome
activity process cycle:
(repeat as many times as needed)
cycle outcome
^ (do) reset | <
| (do) act | < tools
| (do) decide | < artifacts
| (plan) talk | < distal object
| (plan) think | <
| (plan) get info V <
subject(s) - - - - - - - - - - - - - -
rules, conventions community division of labor
_______________________________________
diagram of proposed plan-do activity process model
********************************************************
Thank you for reading this,
- Steve Gabosch
Appendix. The author (me) claims (in truth, tentatively) that this plan-do
activity process model can identify deliverables for both a sign-mediated
decision and a tool-mediated outcome in any human activity cycle. Below is
a summary of the relevant activity cycles in the Wells paper, Table 1. Is
this claim reasonable just for this empirical data in this paper - let
alone all human activity? Can it really work for analyzing all 4th grade
group construction projects not to mention building 21st century skyscrapers?
summary of the main activity cycles:
(the indentations suggest but can't alone demarcate all nestings)
A. 01 - 02. Start of discussion on what to make.
B. 03 - 13 Janet reads her writing to Linda while self-correcting.
Linda listens and writes.
C. 14 - 15 Conclusion of Janet's report ...
D. 16 - 21 Linda works on writing, J negotiates use of book with
C, then looks at book.
E. 22 - 23 Log roller proposal ...
F. 24 - 30 J and L discuss log roller proposal.
G. 31 - 32 Mutual decision to look further ...
H. 33 - 35 General discussion about project while Janet
turns to next page
I. 36 - 37 Girls see and react to sailboat in the book ...
J. 38 - 44 We could use wind power discussion ...
K. 45 - 48 How to make wind power discussion ...
L. 49 - 60 Observations of others, interactions with
others, writing.
M. 61 - 65 Definite decision is made to do the boat ...
N. 66 - 71(end) more discussion of boat and writing
<end>
This archive was generated by hypermail 2b29 : Fri Nov 01 2002 - 01:00:07 PST