[Xmca-l] Units of analysis

[David Dirlam]

It looks like I finally got my blank-emails problem fixed, thanks to Andy's
kind notification of what was the cause. I did a little off-line
corresponding with him in the meantime and plan to share some of it and add
some to it.

I found Andy's current and 2010 articles on Vygotsky's unit of analysis to
be fascinating and his comments during my xmca blackout very helpful, even
after spending 50 years work on the problem. I hope to join this discussion
of what I believe is one of the most central social science problems of our
age. So below I've put a brief synopsis of what I think people on xmca
might find most useful from that work and invite comments.

One of Andy's useful comments about units came through an xmca in response
to David Kellogg's suggestions. The introduction of features (father's job,
sibling number, etc.) reminded me of the attempts to evaluate higher
education in the U.S. using graduation rates and job placements, a tendency
that I have been fighting for decades.

Also, during my offline discussion with Andy, I mentioned a chapter from my
book that used biology as an analogy to define 11 levels of a nested
hierarchy of practices. That was a project that I had been imagining for a
decade by tagging articles in Science that I thought would be useful for
the purpose, especially in the context of many additional articles
generated from library databases (though I wrote this chapter not even a
year ago, it has had much value in helping to identify and talk about what
people do). The 11 levels ranged from an analog to the gene at the bottom,
which was quite similar to Vygotsky's sign-mediated action, since it
combined activity with artifacts and social context. The next level up (the
cell analog), I called actuations, which add short-term memory to the
bottom level, Luria's *Human Brain and Psychological Process *has many
examples of both levels. Procedures that use several actuations (like
recipes) were the next level. The units continue in similar nested fashion
up to the biosphere analog, that I called praxosphere.

Andy replied that units do not have to be nested, and I fully agree. In
fact nearly all the units I discussed have both defectology and social
situation aspects. Since they are not nested like the ones in the chapter,
they need to be different sorts of units. The same occurs in biology where
competition and pathology occur at least at levels ranging from cells to
biomes. Actually, Luria's book just mentioned would be a great place to
identify examples of the first two or three levels of the hierarchy of
practice I proposed in my book. His amazing observations of simple acts of
brain-damaged patients have been an inspiration to me for decades.

The mid-level units (the analog to biology's species) that I proposed are
modes of practice. I found these by studying dimensions of competing modes
of practices in children's drawing, student writing, and developmental
researchers' methods. The first publication of a dynamic analysis of the
drawing study was in *Mind Culture and Activity *in 1997. My language has
changed during two decades of use of the ideas, but the kernels were there.
Dimension is the next more complex level of units above the modes of
practice. The modes of practice in each dimension get sequenced by four
parameters: their endemicity (initial prevalence), acquisition rate (growth
in frequency), and commitment (competitive strength), and their resource
level (social acceptance or limited artifact availability). One pattern is
especially common: *beginning* modes are endemic, *exploring *modes are
acquired very quickly but are not competitive, *sustaining *modes grow
slower with more commitment, and *inspiring *modes have the highest levels
of commitment. Sometimes a *destructive *mode appears instead of the
sustaining or inspiring mode -- these grow faster than exploration, but
overshoot the resource level so much that they eliminate the whole
dimension (drug use is a good example for persons). To transition from one
mode of practice to the next requires transformative learning, a concept
that has benefited much from Mezirow's writings. A study with some friends
of 500 hour long sessions with individual students revealed that Mezirow's
10 phases occurred in 4 time periods. The phases of commitment and modes of
commitment, therefore, became the fourth and fifth levels.

An interesting aspect of the sequence for modes of practice is that it lays
out zones of proximal development for each dimension. I watched teachers
use them that way 40 years ago with amazing results, especially for student
writing. I've had a harder time getting academics to do so. It makes me
wonder how Vygotsky's discussion of zone of proximal development might fit
with the modes and phases of commitment.

We can tell when we have two different dimensions, because all modes of
practice within one dimension can occur simultaneously with any mode of
practice in all others (i.e., there is no competition between them). So,
when I have given this model to experts (300  altogether so far) and asked
them what people do who are learning in their area of expertise, over 99%
readily describe the four or five modes of practice in 6-12 dimensions of
their expertise. They often comment that it is an interesting way to
organize their thinking about their field. My interest is that each
interview results in somewhere between 15,000 (6 dimensions) and a quarter
billion (12 dimensions) patterns of practice but uses only 24-48 terms.
When I did this for the entire faculty of a liberal arts college and
combined dimensions that were similar, I ended up with 25 dimensions with
100 terms (these are detailed in the appendix of my book). They are
incredibly more interesting and meaningful than graduation and
job-placement rates. But, and this is what keeps me at it, they also have
the potential to emancipate teachers and learners from the sort of
bureaucratic hegemony that demands simple minded measures like job
placement and graduation rates.

Text analysis of the 25 dimensions (a technique I started working on at
Mike's suggestion during my 1997-98 year at LCHC) also grouped dimensions
together into clusters that resembled specialties (analog to biological
communities) and these into disciplines (the analog to ecosystems). The
next levels came from interviews of some 80 designers in 20 different
disciplines at the Savannah College of Art Design. I never thought of
design as a methodology as progressive as science, but those interviews
convinced me. Now, it seems obvious: scientists record their progress in
papers and equipment, designers in services and products. Interviews of 60
rabbinical scholars at Hebrew Union College then convinced me that
interpretation was also progressive, but now recorded in precedents. So the
level above discipline (the analog to biome) became progressive
methodology. Human knowledge, then, becomes the praxosphere or analog to
biosphere.

I'm not sure what Vygotsky would think about all this (I am sure that many
on this list know better than me about that), but I don't think he would be
nearly as upset as he seems to have been with reducing a child's
environment to parent occupation, age, housing, sibling numbers, etc.
Thinking about Vygotsky and Andy's papers does give me the idea that I
should look back through all my interviews to see what meaningful objects
and social environments they imply.

I hope this adds to the discussion, and even more that it proves useful for
progress in studying what people do in ways that thwart simplistic
reductions to graduation and employment rates. I'm off to read Yrjo's paper
next.

All the best,

David Dirlam