What is "essentialism," and how should we avoid it?

[Philip Carl SALZMAN]

Whatever our many disagreements on a variety of topics, we can all agree that “essentialism” is bad. We frequently say so. But what exactly is meant by “essentialism”?

My old friend Michael Herzfeld (in his article on “Essentialism” in The Encyclopedia of Social and Cultural Anthropology, Alan Barnard & Jonathan Spencer, eds., which can be found at http://www.bookrags.com/tandf/essentialism-8-tf/) says
Essentialism commonly appears as both a violation of anthropological relativism and one of the besetting conceptual sins of anthropology. ... The distinctive mark of essentialism ... lies in its suppression of temporality: it assumes or attributes an unchanging, primordial ontology to what are the historically contingent products of human or other forms of agency. It is thus also a denial of the relevance of agency itself.

Following Herzfeld, we could say that “essentialism” is a form of generalization or characterization that assumes an unchanging nature unaffected by human action.

What would count as an example of “essentialism”? Let us consider Clifford Geertz’s characterizations of Indonesian and Moroccan cultures (in Islam Observed) as respectively synthesizing and agonistic. These characterizations are certainly generalizations, but do they ignore historical change or human agency? Well, much of Geertz’s account is historical, showing changes over time, and illustrating with the actions and influences of important individuals. So I do not think it would be accurate, using Herzfeld’s criteria, to call Geertz’s general characterizations an example of essentialism.

Another criterion for essentialism, this time put forward by Edward Said (in Orientalism, itself notoriously open to the accusation of essentialism), is the ignoring of internal variation. In Islam Observed, Geertz does discuss variation and some of the reasons for it. So I think Geertz can be acquitted of essentialism in this work.

It is important to note here that avoiding essentialism does not require avoiding generalization and characterization. If an appreciation of the historical nature of the features discussed and the ways in which human agency contribute to the historical pattern are taken into account in the generalization, or in the explanation of the generalization, then no accusation of essentialism is appropriate.

But since generalizing is dangerous, can we avoid it altogether? I would suggest we cannot, because all knowledge is based on generalization. Every individual, whether animal, vegetable, or mineral is unique at every moment of its existence. We cannot even speak of a particular plant or animal without generalizing over time. When we discuss anything beyond the unique, we must generalize based on abstraction of common characteristics. For example, most of our common concepts–tree, car, person, animal, house, rock, etc.–generalize certain characteristics that are common in the class. General though these concepts may be, they are valid: we know that trees in their many varieties, are not foxes in their varieties. My Baluchi friends taught my about the many varieties of camel, but they still spoke generally of camels, which were clearly differentiated from sheep and goats, never mind tents and trucks. Not withstanding the differences among camels, and among horses, it is valid to say that camels are not horses, and that we can specify the general differences between them.

There are other cases where generalizations are based on central tendencies (mean, mode, median), even where the distributions of the cases involve overlap. For example, we know that there are among men and women many heights, and that there are many women taller than many men. But in general (certainly no such claim for myself), men are taller than women. Different groups will give different answers to questionnaires or perform differently on tests (such as Robert LeVine’s study of Nigerian groups in Dreams and Deeds), and in spite of overlapping distributions will show different averages. So too with many measures of social and economic development–e.g. education, health, productivity, connectivity–where different communities, regions, or societies have different results on the appropriate measures(e.g. UNDP Arab Human Development Report 2002).

What of qualitative generalizations, such as those of Geertz? I do not see how they can be avoided if we are to speak about different cultures and their commonalities and differences, both over time and among cases. Applying the intellectual procedures that are typical of all knowledge would appear to be unavoidable.

And yet, it is apparent that anthropologists today do avoid generalizing. Perhaps the fear of essentialism–and other sins, such as “reductionism” and the “master narrative”–leads us to avoid formulating general characterizations. This is apparent in the discussions here on OAC, where in response to any hint of generalization, discussants retreat into anecdote: “Well, the chef in my Japanese soup restaurant says.....” Yes, we can avoid controversy by avoiding attempts to formulate knowledge, but of what value is a substantiveless anthropology? Has fear made us feeble?

[Philip Carl SALZMAN]

Deniz, Does the statement "Moroccan people are generally proud" make Moroccans "look 'bad'"? I am not sure that Moroccans would feel that way. I don't.

Furthermore, I am not sure that anthropology should decide what it says on the basis of how good it makes the people we are talking about look. You would have to rule out statements like "Most Nazis were cruel to targeted minorities," and "Many Soviet citizens were starved, executed, or tortured."

As John suggests, essentialism is "the assertion of uniform, constant and immutable traits that define the culture/group in question," whether those traits might be deemed by some as positive, negative, or neutral. On the other hand, generalizations--descriptive generalizations--can and should take into account internal variation, and should not rule out change.

[John McCreery]

NIKOS, could you elaborate a bit? What do you mean by "HOW TO DISTINGUISH THESE PARTS A PRIORI "?

I ask because, speaking as a pragmatist, I am willing to admit that every inquiry begins with assumptions implicit in the way that the question is framed. I also agree that it is useful to be aware of those assumptions in examining the evidence brought to bear on the question. But this is a very weak sense of "a priori," since the evidence, especially if pursued systematically with sufficient zeal and rigor, may overturn the initial assumptions and require reframing the question.

An odd, but perhaps pertinent, example. Yesterday my wife and I took our grandchildren to see the dinosaur bones at Harvard's museum of natural history. The most spectacular of the exhibits was a forty foot long Pleiosaur, which I learned from reading information provided on a card for visitors with a more than casual interest is not, technically speaking, a dinosaur at all. According to the information on the card, the scientific meaning of "dinosaur" is restricted to reptiles who lived on land and stood erect on two or four limbs that extended straight down from the hips or hips and shoulders. The Pleiosaur lived in the sea and, like many other reptiles (including modern crocodiles and iguanas) had limbs that extended out from the side, It was, thus, not a dinosaur. The assumptions of my youth, in which Pleiosaurs were conceived as dinosaurs living in the sea were invalidated. The bones still exist, the exhibit remains spectacular, but now I have learned something new.

NIKOS GOUSGOUNIS said:

So, you suggest that some parts of generalizations are useful and positive , but some parts are not ( these that don't regard internal variation , shift and whatever change). The point is HOW TO DISTINGUISH THESE PARTS A PRIORI ?

[Philip Carl SALZMAN]

Nikos, statistical measures of central tendency and distribution automatically take into account variation. Qualitative formulations would ideally provide the evidentiary basis, which would almost always include indication of variation.

NIKOS GOUSGOUNIS said:

Philip

So, you suggest that some parts of generalizations are useful and positive , but some parts are not ( these that don't regard internal variation , shift and whatever change). The point is HOW TO DISTINGUISH THESE PARTS A PRIORI ?

[Philip Carl SALZMAN]

Deniz, let me give you a few examples from my own research: About the Baluch, I would generalize as follows:
Sarhadi Baluch are a tribal people. Most Baluch, and almost all female Baluch, are illiterate. Traditionally Baluch were nomads, living in tents and migrating regularly.

Some people might regard being tribal, nomadic, and illiterate as bad. Yet these are accurate descriptors of the Baluch. (You could refer to my book, Black Tents of Baluchistan.)

I could also say about the Sardinians I know the following: Highland Sardinian shepherds look to vendetta to regulate life in the pastures. Invidia, envy, is a sentiment directed at those who strive to do better than others and threaten community unity, and is often expressed in violent attacks. (See The Anthropology of Real Life.)

Vendetta, envy, and violence would be considered bad by many. Some Sardinians think these things are bad. So we should not describe them, should pretend they do not exist?

If we anthropologists are trying to study human life in all of its forms, its commonalities and its variations, would we not have to describe life in its various aspects, whatever we or someone else thinks of them?

[deniz batum]

Philip, the statement "Moroccan people are generally proud" was from a Dutch writing on Moroccan culture (I read it a long time ago so I do not remember the name etc, but I remember its arrogance), which degraded Moroccan people for being proud, irrational people vis a vis the rational Dutch. So in this case, I do think it is essentialism.

As John suggests, essentialism is "the assertion of uniform, constant and immutable traits that define the culture/group in question," whether those traits might be deemed by some as positive, negative, or neutral. > Yes I agree with that.

Philip Carl SALZMAN said:

Deniz, Does the statement "Moroccan people are generally proud" make Moroccans "look 'bad'"? I am not sure that Moroccans would feel that way. I don't.

Furthermore, I am not sure that anthropology should decide what it says on the basis of how good it makes the people we are talking about look. You would have to rule out statements like "Most Nazis were cruel to targeted minorities," and "Many Soviet citizens were starved, executed, or tortured."

As John suggests, essentialism is "the assertion of uniform, constant and immutable traits that define the culture/group in question," whether those traits might be deemed by some as positive, negative, or neutral. On the other hand, generalizations--descriptive generalizations--can and should take into account internal variation, and should not rule out change.

[Philip Carl SALZMAN]

Deniz, so it is not the statement, but the general political view that should define "essentialism"? Anthropology is now to be defined by political views, some correct and others incorrect?

You have not dealt with the other examples I give in that post, or the examples from my own work that I give in the following post.

[deniz batum]

Dear Professor
Everything everywhere is always defined by political views. Don't you think so?

Philip Carl SALZMAN said:

Deniz, so it is not the statement, but the general political view that should define "essentialism"? Anthropology is now to be defined by political views, some correct and others incorrect?

You have not dealt with the other examples I give in that post, or the examples from my own work that I give in the following post.

[deniz batum]

Hello

I completely agree with your claim of anthropology being a social construct. I want to congratulate you for this. Anthropologic knowledge is always situated, always a construct. I do not understand why people can not see that, or whats so wrong with saying that. Perhaps us the younger generation of anthropologists, have been subject to too much post structuralism.

BDwyer said:

Hello again, my apologies about the long delay in responding, been busy on other things then lost my password which i just found. Anyway, this reply is directed towards Salzman and McCreery I guess and in response to this my reference to "absolutist science". So if you don't want to pick the thread up again that is fine but i thought it good manners to respond. By absolutist science I mean precisely what you are talking about - this sense that ethnographers feel what they do is always must approach being essentialist without going all the way. By absolutist I mean that is how they frame their world view if you like - that there is a mythical unity binding us all and we social anthropologists go about delineating with ever so careful nuance the subtle variations and emergent distinctions. I'm fine with that as I argue that all knowledge is highly contingent and rightly so - there are some facts about the world to be sure, but most stuff to do with human beings is messy and contingent. But this means any necessary unity is a heuristic only. To ask the question posed as it was here is to reify this unity as a real object or even an ideal. I wanted to point out that social and cultural anthropology is a social construction which simply means that social anthropology cannot be thought of as dealing with facts - as in there is a tree there or not there; that social anthropology is intended action and that we apply social anthropology to the world and we are not that clear on why we take something as evidence. I guess this is why the anthropologies have shifted more firmly into the humanities in recent years.

What do you think?

thanks

[deniz batum]

Consider, for example, the data I used for my dissertation on the symbolism of popular Daoist magic. The stuff about which I was writing, the spirit money, the incense, the food offerings, the mudras (mystical gestures) and the words employed in incantations are as thoroughly material facts as trees.

Just because you are writing about "material facts" does not mean your production is not socially constructed. And furthermore may I ask who has conceptualized this "magic"? Anyway, you are writing about the implications, meanings of material facts. You are writing this to share with an audience for whom it will make sense when they read it. You use a certain "language" so to speak to convey this information. If all of this is not socially constructed I do not know what is.

John McCreery said:

BDwyer writes,

I wanted to point out that social and cultural anthropology is a social construction which simply means that social anthropology cannot be thought of as dealing with facts - as in there is a tree there or not there

I reply,

This is nonsense. Consider, for example, the data I used for my dissertation on the symbolism of popular Daoist magic. The stuff about which I was writing, the spirit money, the incense, the food offerings, the mudras (mystical gestures) and the words employed in incantations are as thoroughly material facts as trees. The same is true of the credits attached to winning ads in the Tokyo Copywriters Club Annual that I am analyzing using the techniques of social network analysis, which, by the way, are mathematically well-defined and produce both predictable and replicable results.

The truth in the assertion that science and scholarship are social constructions is that agreement on what counts as a plausible conjecture or solid result depends on building a consensus within research communities that share basic concepts and methods, which, by the way, need not be the same from one community to another.

As Aristotle points out in Nichomachean Ethics III, it is the mark of an educated person that they demand no greater precision in argumentation that the subject matter allows. Literary scholars looking for fresh insights into particular works or genres confront different kinds of data and employ different interpretive approaches from biochemists performing experiments to analyze the structure of proteins. Mathematical logicians demand more rigor than archeologists developing plausible conjectures from the spatial distribution of potsherds.

There is nothing, however, in this scholarly common sense that implies that social construction equals sloppy all the way down.

[Philip Carl SALZMAN]

Certainly not. If this were true, then we can forget about research altogether, as we all have our answers a priori. The view that everything is political and can be nothing else, and the celebration of it, is the most corrupting influence in anthropology. This view destroys any idea of knowledge, and makes anthropology no more than an opportunity for the expression of personal and political impulses, and thus worthless.

deniz batum said:

Dear Professor
Everything everywhere is always defined by political views. Don't you think so?

Philip Carl SALZMAN said:

Deniz, so it is not the statement, but the general political view that should define "essentialism"? Anthropology is now to be defined by political views, some correct and others incorrect?

You have not dealt with the other examples I give in that post, or the examples from my own work that I give in the following post.

[John McCreery]

Deniz Batum writes,

Just because you are writing about "material facts" does not mean your production is not socially constructed. And furthermore may I ask who has conceptualized this "magic"? Anyway, you are writing about the implications, meanings of material facts. You are writing this to share with an audience for whom it will make sense when they read it. You use a certain "language" so to speak to convey this information. If all of this is not socially constructed I do not know what is.

Deniz,

Like Phil, I ask if you read beyond the first paragraph to which you react with current clichés. If you had, you would have seen me saying directly that scholarly knowledge is socially constructed. But for me this is just the start of a discussion that includes close attention to how it is constructed and the implications of different forms of construction. Personally, I agree with Pierre Bourdieu's position that is perfectly possible to acknowledge that the canons of scientific method are a product of particular places and moments in history and to, nonetheless, to find them worth defending. They have, after all, produced results like the technology that we are using to communicate; a claim that no other approach, religious, aesthetic or common sense can equal.

Note, too, the quote from the Nichomachean Ethics. I am under no illusion that anthropological fieldwork can produce results comparable to those of experimental science. How could it, since most fieldwork is done by individuals with different degrees of talent and training, pursuing what insight they may discover while deploying ideas and methods that are little more than starting points and rules of thumb to work with, producing work that is rarely if ever replicated? But that does not change the force of what I was saying when referring to material facts about the body of ritual I studied for my dissertation or the ad annual credits that supply the primary data for my current research. There are facts there, facts firmly attested and readily available for other researchers to examine; facts relevant to judgments that this or that interpretation is better than others. The line you are pursuing leads nowhere but to Punch-and-Judy shows in which "critics" with different prejudices talk past each other. The word for that is piffle. Not my cup of tea.

[deniz batum]

I do not agree.

Philip Carl SALZMAN said:

Certainly not. If this were true, then we can forget about research altogether, as we all have our answers a priori. The view that everything is political and can be nothing else, and the celebration of it, is the most corrupting influence in anthropology. This view destroys any idea of knowledge, and makes anthropology no more than an opportunity for the expression of personal and political impulses, and thus worthless.

deniz batum said:

Dear Professor
Everything everywhere is always defined by political views. Don't you think so?

Philip Carl SALZMAN said:

Deniz, so it is not the statement, but the general political view that should define "essentialism"? Anthropology is now to be defined by political views, some correct and others incorrect?

You have not dealt with the other examples I give in that post, or the examples from my own work that I give in the following post.