Geertz's first chapter in his collection of case studies is, among
several other things, a fascinating introduction to Gilbert Ryle's idea
of "thick description" as the ethnographer's objective. He presents
thick description as an archetype of an interpretive, semiotic
philosophy which states that "culture consists of socially established
structures of meaning." The contribution of thick description is in
providing people with rich material with which to think about and think
"with," when pondering or tackling complex problems.
However, I'm baffled by Geertz's vigorous distancing of his approach
from that of cognitive anthropology. And when he started to sound like
Sancho Panza, throwing clever aphorisms into every other sentence, I
found it best to skip the parantheticals (at least this first time
through the chapter). On one hand Geertz asserts interpretive theory
as being based on a "semiotic concept of culture" while at the same
time he bashes Goodenough's cognitivist approach as engendering such
dreadful "ingenuities" as taxonomies, tables, and trees. These are
certainly semiotic constructs for understanding how things work and
what they mean. I must not be reading Geertz's argument correctly, or
perhaps the points of divergence are a matter of degree; in any case
it's a good idea to keep out of family arguments. For my own part,
ways I would need to think about things in order to achieve a thick
description will have to include cognitive, semiotic, phenomenological,
deconstructivist, and whatever other approaches I could bring to hand.
I liked the idea of the ethnographer as somebody who has to first grasp
and then render socially established structures of meaning. This is
certainly complex, layered, "thick" work. I also liked the focus on
the ethnographer's job as being to "inscribe social discourse," turning
it from things that happened into an crafted account that enables
others to have insight into the thing described.
Geertz's characteristics of ethnographic description: it's
interpretive; it interprets the flow of social discourse; it attempts
to extract meaning from this discourse (rather than simply list
events); and it's very focused, extrapolating larger pictures from
"extended acquaintences with extremely small matters." Getting from
the focused miniature to the wall-sized picture is a problem that is
also relevant for non-anthropologists who want to draw from the
discipline of anthropology.
The section on the difficulty of developing theory is also interesting.
Without clear formulations of theory it is difficult to compare and
assess the results of one's efforts (e.g. one's interpretations). If
the purpose of theory is to develop a vocabulary with which the
objective of the discipline can be expressed and achieved, then there
is certainly a need to develop a theory of information architecture.
That ethnographers are now found on commercial organizational change
teams and commercial marketing projects, and that disciplines like
information architecture are looking to anthropology for theoretical
and practical guidance, shows that anthropology's "grander ambitions"
are to some extent being realized.
Like anthropology, the nature of IA as a discipline and the state of
its philosophy are such that we continue to work and rework the ground
we have (in fact) only just begun to till. This is well-expressed in
Geertz's statement that "[s]tudies do build on other studies, not in
the sense that they take up where the others leave off, but in the
sense that, better informed and better conceptualized, they plunge more
deeply into the same things."
>Geertz seems to really be a writer more than anything?
Yes I think he does self-identify as a writer (more specifically a
rhetorician). According to an interview from 1990 (attached) it's all
about the writing; he took philosophy as an undergrad, and he has an
unpublished novel or two from his youth (although who doesn't ;))
>He talks about his work in terms of what he's written, as opposed to
"who he's studied" >or "what he's done" - I guess if you take the
notion that this is about observing stuff >and WRITING IT DOWN, then
that is a frame of reference you bring to your >professional identity.
Right. Stuff has to be written down in order for it to be shared with
other people. That's so obvious a statement, that it can be
taken-for-granted - what we see is what we write - and many
anthropological accounts imply that what they are writing/presenting,
is an accurate 'picture' of reality. So one of the things that Geertz
questions - and this was much more of a revelation in the 1970s - is
the picturing relationship between what is written, and what is
According to (my interpretation) of Geertz, there is often an implicit
assumption that what is written in ethnographic accounts, reflects
what actually occurs/occurred. (Although this assumption is less
implicit today than it was in Geertz's time.) This assumption is not
necessarily asserted in each account; it's just accepted as fact, for
various reasons (historical, cultural, philosophical) by those who do
ethnography and anthropology. However, consider the following.
1. How field work is carried out by anthropologists. You pick a field
site, you attempt to learn some of the language, you arrive and find
informants, you start collecting data; then you go home and write up
your notes. This process involves a number of interpretive leaps. You
don't know the language perfectly, or the dialects; your informer is
not a neutral observer with perfect information, but situated in the
local culture (perhaps s/he makes their living from anthropologists
passing through); your questions and notes are selective (and you
don't write everything down anyway, and what you do write down may be
imperfectly recorded); when you get back you sift your notes,
order/select from them, interpret them *again,* assemble them into
texts, and so on.
As Geertz says, ethnographic accounts are interpretations of
interpretations of interpretations, to the nth order.
2. What these interpretations, interpretations of? Geertz argues
against the idea that culture is a system/collection of defined
variables, which people have to learn about to acquire that culture.
Things like kinship structures, property ownership patterns, and so
on. Rather, he says that culture is shared between people, in the
things they do, and the symbols they use while they are doing whatever
they are doing.
I guess this sounds a bit vague, so let's try another way. One way to
see culture (which Geertz opposes) is as a collection of cultural
knowledge, which we acquire over the course of our lives. How to talk
to someone, how to behave as a member of a family, how to buy
something, how to drive a car, and so on. In this model, we acquire
cultural knowledge a bit like Legos, and assemble them into an
internal model of the society we live in.
Individual cultural Legos are however not things in themselves. They
are always defined in terms of other cultural facts - some of which we
may have access to, others of which may be new to us. So we don't
acquire them sequentially, but rather concurrently, and not as
predefined units, but rather in the context of however it is they are
(culturally) defined at the time (e.g. a wink as twitch, wink,
conspiracy, parody, practice, etc.).
(1) and (2) are obvious claims for cultural fluidity and relativism,
and depictions of culture as a fluid and relative phenomena. So how do
we go about analysing this? One answer is thick description,
approaching the same phenomena from multiple (interpretive) angles.
Anthropology is as much writing work, as it is observing work.
>Later, he makes reference to two models "Jonesville-is-theUSA
'microcosmic'" and >"Easter-Island-is-a-testing-case 'natural
experiment' - what is this about? The >references are lost on me.
I think here he is referring to:
(a) Looking at the 'average' American small town as a lab model of
American society as a whole;
(b) Looking at isolated/unique cultures - again as lab models.
Thanks to Steve for starting us off!
Here's my 2 cents.
This notion of "operationalizing" I believe to be a reaction to other
researchers working in Geertz's time. In a sense, he's swimming upstream
against a "scientific paradigm" (to cite Kuhn for a sec) that the only
social research worth doing is one in which concepts are
"operationalized" into variables.
This is not a positivist science, he argues. Now haven't we all faced
that, from, say, a usability perspective? What does it really mean that
a 41% of users completed this task? Wouldn't it be more interesting to
learn, for example,
"'Hugh' sat behind the keyboard and glanced hesitantly at the screen.
'Is this a 12 inch? I have a 15 inch at home.' The moderator, standing
beside the screen and facing Hugh, asked him first what his reaction was
to the home page. 'What? This is the home page?' he says, squinting. 'It
looks like something else...like a category page.'"
My interest in this work stems mostly from this quote:
-- "man (sic) is an animal suspended in webs of significance that he
himself has spun"
In other words, deep insight into (user) behavior requires understanding
the symbolic universe. We can only do so from observation PLUS
The interpretive tradition is at odds with much of the history of
software or technology design; their bases in scientific positivism
leads naturally to quantitative versions of validity. Hence the
meaningless findings of task completion. I want to know what people
think, how they react, how or if they "wink."
So cultural knowledge is a system of interrelated symbols (the wink, for
example, must be viewed within the context of when it happens).
K., now to take it crazy crazy for a second (bear with me), I have
questions about the interpretive tradition vs. the phenomenological
tradition. The "vershten" (read: Weberian) approach of interpretation,
Geertz admits, often leads to the idea of long-distance "mind reading."
But his response that interpretation is NOT simply mind-reading is
inadequate. For interpretive researchers, meaning precedes
interpretation. But for phenomenological researchers, interpretation IS
meaning making in itself.
So what this means is, watching a person as they use their cell phone at
a bus stop, by its very act is creating meaning. There was no meaning
before we stopped to observe her. I think others critique Geertz and his
ilk based on this. The entire approach to observation is, in itself, the
creation of meaning.
But I digress. What does this mean for user research? It means that we
must slog through reams of description, decipher symbols (spoken, worn,
expressed, etc) embedded within that description, think really hard
about where those symbols fit with what our original research question
was, and come up with an interpretation (without ever losing sight of
the fact that by the very act of interpreting we are creating meaning
that is distinct and has a Gestell or method of "revealing" that is tied
essentially to our act of investigating).
The information contained in this message is confidential. It is intended to be read only by the individual or entity named above or their designee. If the reader of this message is not the intended recipient, you are hereby notified that any distribution of this message, in any form, is strictly prohibited. If you have received this message in error, please immediately notify the sender and delete or destroy any copy of this message.
I'm afraid that Friday was a little optimistic for me as a deadline.
I know that I am probably unknown to most members of this group. My
name is Byron Burkhalter and I am a PhD in Sociology (UCLA). Most of
my work has been in ethnography, phenomenomology, and ethnomethodology.
I have done some contract work with Microsoft which focused on using
findings from conversation analysis and ethnomethodology to user
Thanks for bringing me back to Geertz. It's great to read the
honesty, plainspoken criticism of colleagues, and humility that tenure
at your university and stature in your field allow. I enjoyed reading
Here's what I got and what I thought about:
1. Geertz begins with "culture" as the point of anthropology and a
definition of culture as semiotic. To me that means he is on a search
for deeper meaning for social action than mere behavior would offer.
Semiotics has is that actual meaning comes from an understanding of how
each element fits into all other elements in a system. Thus the
ethnographer's job is not simply observing social action or behavior
but interpreting the meaning behind and beyond that behavior. By
making a deeper, hidden meaning the object of his work I think that
Geertz has created a struggle for himself. He recognizes the
importance of analyzing actual interaction but he also wants to get
beyond that to find what he thinks is significant. That makes the
interaction less important than interpreting the meaning behind it.
2. The second section demonstrates a piece of thick description
although I could not distinguish it from "telling a story." After
the excerpt Geertz makes two claims. First the claim that background
understandings help one understand what is going on. I agree only I
would focus on participants' use these background understandings, not
so much that anthropologists have to find these in understanding the
events they record. Second Geertz points out that we as ethnographers
are faced with "a multiplicity of complex conceptual structures."
I think this is misleading, I think we are faced with what we see and
that is people in interaction. Lastly, the final part of the last
paragraph of this section is absolutely how I see my own field
notes-particularly the incoherent comment.
3. The third section begins with a claim that culture is public. This
seems to contradict the notion that culture is semiotic. If culture is
behavior then it is public but if cultures signifies some other import,
that other import is not public. If you need to use interpretation to
get to the import then I take it that import is not socially available
outside some mental, and thus hidden, process. His solution to this
seems likely enough. He sees culture as having the cognitive process
and the behavior intertwined.
4. On page 11 he starts to identify what the work should not do. He
seems particularly cogent and effective in this task. He doesn't
want the work to reify some metaphysics, so you can't bring in
cognition or capitalism or other external forces to explain a
particular cultural configuration. But without the psychological or
sociological motivations to explain culture one would seem left with
mere behavior and he clearly doesn't want that outcome (and I agree).
He correctly sees the problem with psychologism to be giving everything
to something inside the subject's head that we have no direct access
to and in fact that is not empirical. Therefore he goes again to the
statement that "culture is public because meaning is." Okay then
where is meaning. If cultural meaning is not in the head and not in
massive social forces like capitalism and it is public then where is it
to be found? Geertz says (p.15) that meaning is in the imaginative
universe. Okay, I give, where is that?
5. The 4th section makes the point that ethnographic work is
necessarily second or third order and that ethnography doesn't bring
real culture back in its articles. And now he begins to give ground.
With his view of interpretation I think he has written himself into a
6. In section 5 Geertz goes completely against the structuralism of
which I claim he is a member and which founds the first part of this
paper. The first paragraph is a nice explanation of structuralism. And
this brings a much nicer definition of what I hope my work is about:
"the informal logic of actual life."
7. Now he has discounted cognitive psychology, grand sociology,
structuralism and behaviorism. That means that culture is more than
what you see in observing behavior. The "more" of culture is not
however in the mind, social forces or the relationship of elements in a
system. Toward the end of the section he offers that cultural analysis
is a series of guesses but doesn't say from what source educates his
guesses or what standard he uses to assess the guesses of others. He
seems to have run out of rigorous solutions.
8. Section 6 is for me a compelling argument for researchers to stay
close to the actual events they study and that claims must have actual
instances to ground them. In the final two sections only three things
stood out for me: (1) I like the idea of theory staying close to the
ground of actual events, (2) the inability to use this method for
prediction may be accurate but, if true, has some aspect for corporate
work and (3) the end of the piece has Geertz seeing the value of his
work in the transcribing of other cultural methods not so much in the
interpretation of those methods.
What I probably did not make enough of was how much I like Geertz'
argument that the process of ethnographic research looks messy,
disorderly and haphazard. I look at my own notes (and thoughts above)
and know that it is all too true of me. That's okay, I like the
mess-to a point.
I love this, Stacy. Do we think that the "family argument" issue is
what would keep Geertz, at least at the time he wrote that, from
agreeing with you in your inclusive approach to approaches? One might
(inaccurately) accuse you of being kitchen-sinky about it, and I
think that would put Geertz off, but I bet really you are simply
bringing in multiple perspectives to help you keep honest with your
"thickness" and I think THAT is right inline with Geertz.
Steve Portigal -- http://www.portigal.com
I wonder what Geertz would make of the video-camera-revolution.
There's a pseudo-anthropology in reality TV, in the aesthetics of TV
advertising, of YouTube, and viral video. But even within
ethnography, the electronic data gathering of video (and to a lesser
extent audio) as a proxy-for-truth - the camera is operated by a
person who is an implicit editor, but the interpretation in the
moment is granted less than within a pen and paper.
The technology has *changed* the practice of writing. We can
transcribe video, or we can bring in a laptop and transcribe while
it's happening, or we can make notes from a video record rather than
from the event itself. I know I'm talking about process here, but my
point is more connected to the theory of the role of writing - how do
we evolve that theory now that the tools have changed and new
behaviors are enabled?
Steve Portigal -- http://www.portigal.com
Would you throw the word "bias" in there?
If we are creating meaning when we stop to observe something, was it
meaningless beforehand? I guess this is a tree-falls-redux, so sorry
to open a whole thing here if we needn't :)
Could one make an argument that the meaning created is a win-win form
of interpretation, for if we are as you suggest focused on our
research problem - our actual business/design agenda, the meaning we
are creating could potentially be inherently tied to what we're
thinking about anyway, what we're tasked with?
Straw grasping, perhaps....
Yes, totally this is a tree falling in the forest idea. Need we make
this? Hmm. No. Could we, should we? Why not?
Quantum physics tells us that the very act of observing determines
outcome. Is that not important to remember for user research? Of course!
There are degrees of determination (e.g., contrived usability lab
setting vs. ethnographic in the field observation) but treating "users"
as an object of study, we are creating a certain approach to user
Instead we ought to be walking through the mall, on our way to dentists'
appointments, noticing a group of exchange students holding up cell
phones and a coupon flyer for the pharmacy next door, and trying to get
the cheap price, as well as trying to give each other their cell phone
numbers....and this observation is unintentional. It was not engineered
or "measurement" in any sense.
What does such an observation mean for user experience design? We could
come up with dozens of ideas between mobile marketing, GPS location
systems, language dictionary cell-phone software, etc.
The method of observation determines how the world is revealed to us.
> Could one make an argument that the meaning created is a win-win form
> of interpretation, for if we are as you suggest focused on our
> research problem - our actual business/design agenda, the meaning we
> are creating could potentially be inherently tied to what we're
> thinking about anyway, what we're tasked with?
> Straw grasping, perhaps....
> Steve Portigal -- http://www.portigal.com
> Note: blog at http://chittahchattah.blogspot.com moving soon
> to http://www.portigal.com
A potential 'problem' with this I guess - besides having to become
skilled in anthropology *and* filmmaking - is that the community of
practice in which this occurs is much smaller, and the body of
videos/texts available is also therefore much smaller. We make sense of
Geertz himself in myriad local ways, for instance, related to other
things we have read/written. But we're skilled in reading, and making
sense of ethnographic film may be a much harder task, as we also have
to learn how to read film, and the visual/rhetorical strategies of
filmmakers in general, and ethnographic filmmakers in particular.
While video is different to writing, but unless you deal with the video
only, for instance in Final Cut, and refrain from writing anything, the
physical process of inscription is not replaced, but only deferred.
That is, you will get around to writing about it sooner or later. And,
video is also just another text, in the postmodern everything-is-a-text
sense. So I guess I would say that video does not necessarily require
radically new theoretical strategies, although it does require film
theoretical strategies of some sort. The theoretical work here,
however, at least from the point of view of social science/humanities
types, is probably more to do with broader philosophical and rhetorical
questions that include writing and film, rather than with any specific
film/video based theory. But then again I was never a film studies
student, maybe they would have something different to say!
I think that the 'process' question is still interesting. At the moment
I do a lot of work in organizational settings, and have used video
extensively in the past. Before that I used to rely on heavy
note-taking. And there was a period in which I did both together, and
it was during that period that comparing my notes with my video, that I
realised exactly how selective/forgetful/threadbare my notes were. Now
I do a lot of digital video, and apart from anything else, it's nice to
be able to digitise the video, rip the audio, ftp the audio file to a
transcription service, and get the transcript back for further
analysis/processing/coding, all in a couple of days.
> Do we think that the "family argument" issue is what would keep
> Geertz, at least at the time he wrote that, from agreeing with you in
> your inclusive approach to approaches? One might (inaccurately)
> accuse you of being kitchen-sinky about it, and I think that would put
> Geertz off, but I bet really you are simply bringing in multiple
> perspectives to help you keep honest with your "thickness" and I think
> THAT is right inline with Geertz.
I think Geertz would see my inclusive approach as dilettantism. And he
would be right. However, though I am a dilettante, I take courage from
the example of Indologist Heinrich Zimmer. He believed strongly in the
importance of having a dilettante attitude towards the subjects we
study. It's humbling when one realizes how very little one knows; but
with this humility comes an open-mindedness that actually brings one
closer to discoveries worth making.
For me, the value of this theoretical discussion is ultimately in how
it can inform my practice. This makes me not only a dilettante but a
> Instead we ought to be walking through the mall, on our way to dentists'
> appointments, noticing a group of exchange students holding up cell
> phones and a coupon flyer for the pharmacy next door, and trying to get
> the cheap price, as well as trying to give each other their cell phone
> numbers....and this observation is unintentional. It was not engineered
> or "measurement" in any sense.
> What does such an observation mean for user experience design? We could
> come up with dozens of ideas between mobile marketing, GPS location
> systems, language dictionary cell-phone software, etc.
> The method of observation determines how the world is revealed to us.
Contextual inquiry is a formal method that lets us partner with users
(rather than just study them) and focuses us on developing useful
products as a result of the research. I have the notion that if
contextual inquiry were more widespread, even the "ordinary" user
research would get better.
Since contextual inquiry is a method or practice, however, is it
outside the scope of what we're discussing right now?
It's pretty interesting (and exciting) to get that argument in
support of mess from a theoretical discussion, where theory in
general could be thought of as bringing order to things, rather than
celebrating or at least acknowledging the necessity of mess.
Video may be different than film because of it's plentifulness -
compared to film, video is cheap, unobtrusive to collect, includes
audio, where you can collect hours and hours and hours. There may be
a cultural bias towards video as representing "everything" or
"reality" - or being inherently thick, etc. I guess that was what I
was poking at.
Theoretically, it is a text, but as a technology, it affords
"powerful" uses that may shift it towards supplanting another form of
text, or to simply limit our "writing" because of our beliefs around
I had to look that one up. From Wikipedia "A bricoleur is a person
who creates things from scratch, is creative and resourceful: a
person who collects information and things and then puts them
together in a way that they were not originally designed to do."
Hmm, that's interesting. I think that there's a cultural tension about
whether video represents reality or not. In one sense, the camera is
indeed seen as not lying - for instance the use of 'eyewitness' cameras
on news reports, surveillance cameras, cameras/film/home movies as
souveniers, documentaries, and so on. But then you also have special
effects/CGI/reality shows/fake reality shows and so on. People are well
aware I think of the artificial possibilities of video, and indeed can
like being faked/fooled/gamed, e.g. in special effects. But then there
is this strong sense of 'the camera never lies' in ethnographic film
which can, in its mannerisms, can aim to reproduce the visual and
rhetorical strategies of the documentary/surveillance genres, and
convince us of its truthfulness.
It's an interesting point, that there may be a bias towards seeing
video as inherently thick. I'm not sure whether (raw) video is in
itself inherently thick. There is an awful lot of detail in video -
microethnographers, for instance, use video as their evidence, and
construct accounts of human interaction in which their unit of analysis
can be as small as a tenth of a second, and then write whole
papers/chapters about the significance of 15 seconds of interaction -
but is this detail necessarily thick, or just long?
I really like the documentary work of people like the Maysles brothers.
The Maysles are credited I think with the 'fly on the wall' style of
documentary making, and they appear to present a narrrative arc in
their documentaries, without introducing themselves. But are they
presenting one long description, or a multi-layered, multi-perspective
view - the same tree shot at dawn/midday/dusk? In this sense, I don't
see thick description as necessarily involving *long* description
however; and in this sense I do not see video as being necessarily
thick, or supporting thick description. Thick description for me
involves rather multiple points of view; a series of 'on-the-one-hand's
and 'on-the-other-hand's that builds up a picture of what may or may
not have taken place. Thick description involves revisiting and
criss-crossing the same terrain with an open mind, contextualizing and
recontextualizing, rather than building up a single definite picture
through a long description.
Perhaps this idea that video is inherently thick, arises out of a lack
of reflexivity that different meanings could be ascribed to the same
> Theoretically, it is a text, but as a technology, it affords
> "powerful" uses that may shift it towards supplanting another form of
> text, or to simply limit our "writing" because of our beliefs around
> its power.
I think that we definitely are getting more accustomed to
shooting/watching video, rather than writing ...
Steve Portigal wrote:
> It's pretty interesting (and exciting) to get that argument in
> support of mess from a theoretical discussion, where theory in
> general could be thought of as bringing order to things, rather than
> celebrating or at least acknowledging the necessity of mess.
One of the reasons I distrust most social theory is that it does try to
make things too orderly, not acknowledging the mess of reality. There
are a few branches of theory that do acknowledge such mess, though, and
they make a lot more sense to me. Geertz's thick description is one.
Garfinkel's ethnomethodology is another - check out the way he makes
explicit all of the shared context necessary to have an everyday
conversation at http://www2.pfeiffer.edu/~lridener/courses/ETHNOMET.HTML
I've been reading Bruno Latour, who comes out of the world of science
studies, and completely disdains the idea of social theories, because
they oversimplify the world; one of his quotes about social theories is
“They design a picture which has no gap in it, giving the spectator the
powerful impression of being fully immersed in the real world… it’s this
excess of coherence that gives the illusion away.” I'll try to dig up
an article-length work of his that might be suitable for this forum.
Thanks for organizing the group, Steve, and for being the engine moving
it forward. I'll try to start pulling my weight around here.
I saw this cited as an example of historical thick description; it's an
It's from the book "The Great Cat Massacre : And Other Episodes in
French Cultural History," by Robert Darnton.