Donate Menu

Ruth Benedict's Obituary for Japanese Culture: An Exchange

December 1, 2007
Volume 5 | Issue 12
Article ID 2597
Ruth Benedict’s Obituary for Japanese Culture: an exchange

Toru Uno and C. Douglas Lummis

What is the nature of Japanese Culture? Japan Focus published Douglas Lummis’s critique of Ruth Benedict’s Chrysanthemum and the Sword, arguably the most influential work ever written on Japanese culture.

Below find a response from Toru Uno and Lummis’s rejoinder. Japan Focus welcomes further contributions to this debate.

How to Critique: Lummis on the Legacy of Ruth Benedict

Toru Uno

Ruth Fulton Benedict’s intellectual presence is still being felt in the field of comparative cultures and beyond, even though almost six decades have elapsed since her passing. We have come to see her epistemological orientation as our own as much as uniquely hers. So much so that we are no longer conscious of our indebtedness to her pioneering work today. In observing a culture different from our own, we try, almost instinctively now, to elicit a discernable pattern while being mindful of its own internal logic. This disciplined perspective toward a foreign culture seems even more relevant and indispensable in the often emotional “kulture kampf” debates we encounter in the post 9/11 world.

It is opportune in this context for Japan Focus to call attention back to Benedict’s intellectual legacy by posting a critical essay by C. Douglas Lummis, “Ruth Benedict’s Obituary for Japanese Culture”, on July 19, 2007.

With such a provocative title, it is hard not to notice this essay even in this age of information overload, accelerated by ever-expanding online publications. In his newly added preface, Lummis explains the evolution of his essay since the original version was published in early 1980s and notes this updated piece meant to be his, or probably the definitive statement on Benedict’s intellectual legacy. With all this anticipation building, I was sufficiently intrigued and even eager to read it through at one sitting, even though it is rather a long piece for an online essay, 20 single-spaced pages.

My anticipatory enthusiasm evaporated rather quickly, however. In fact, it was replaced, with equal velocity, by disappointment bordering on a sense of futility. The sources of my discontentment with Lummis’s “logic”, if you call it such, may be summarized broadly into the following categories:

• the essay lacks the methodological discipline necessary to formulate and justify a balanced and evidence-based conclusion. The evidence in the essay, if provided at all, is nowhere close to justifying the sweeping and overstated conclusions;
• the various accusations he makes against Benedict and her work remain conjectures driven by his particular ideological stance. The Russian proverb, nyet faktov, tolko versii (there are no facts, only theories), seems to fit this speculative essay;
• Lummis’s argument seems to have been built on a different epistemological foundation from Benedict’s. While Benedict viewed culture in dynamic, behavioral terms and tried to elicit a pattern, or an (ideal) model, of Japanese culture, Lummis seems to have mistaken her thesis simply as a static, factual description of Japanese culture. This fundamental conceptual discrepancy, combined with his erroneous view on the role of an “ideal model” in a social and behavioral science inquiry, inevitably led to his skewed assessment of Benedict’s legacy.

I. Methodology, Logic and Other Inconveniences

It should be noted at the very outset that it is no easy task to follow Lummis’s argument. It requires a sympathetic and imaginative reading to achieve even a semblance of a logical flow. This should not be misconstrued, however, as condoning the current preoccupation with simplifying a complex issue into a few sound bites and/or an easily digestible, short video clip a la YouTube. Yet, if we are forced to come up with some logical flow of ideas, then something is amiss. Upon reading, it became somewhat easier to appreciate the dilemma his earlier editors had to deal with. They allegedly took the liberty of revising Lummis’s earlier version of the same essay without his knowledge or his consent, as noted in his preface.

Fundamentally, the difficulty stems from the very nature of this paper: it is, to me at least, a disjointed set of opinions, rather than a disciplined argument logically constructed and rigorously tested through cross-referencing with objective evidence. The article leaves the reader not quite sure of what the main points the author really wants to prove, other than the uneasy, yet distinctive, impression left behind, namely his single-minded zeal to direct the thrust of the essay toward the inevitable conclusion that Benedict’s study is “deeply flawed”.

Everyone is and should be entitled to have his or her opinion. Lummis should certainly not be denied enjoying this very privilege. Yet if a paper has a pretense of being an academic discourse, particularly when it is to make grandiose claims as he does, we are entitled to expect that the minimum, generally accepted standard of a logical discourse to be employed in justifying his conclusions. It would have been helpful if an argument began with a clear definition of a problematique, which is then rephrased into a few testable or verifiable statements or hypotheses. Before reaching any conclusion, the rephrased statements should be scrutinized and verified in light of as much data/evidence as one can marshal. All these steps are nothing special; just ordinary, logical and sequential steps found in a basic, introductory textbook of social science methods assigned to entry level college students. A clear definition of key words and their consistent use throughout the essay might have been helpful as well. For instance, Lummis should have provided his usage of the crucial term, culture. Should his usage be different from Benedict’s, it would be entirely possible the whole discussion might have been rendered pointless before it began.

He begins with the following three general statements, which I took as his “working hypotheses”:

• the “popularity” of her work owes singularly to her exceptional skill as a writer [and, by implication, not by substance at all?];
• the success of her work owes to the post war euphoria of surging liberalism in the USA;
• Benedict’s work is “deeply flawed” according to Lummis’s own fundamental criterion in assessing any work on Japanese culture; “whether it helps or hinders understanding of Japanese culture.”

If I followed his paper correctly, an unstated derivative to the last of the “working hypothesis” above may be as follows;

• Benedict (allegedly) relied exclusively on an informant who is not authoritative, nor reliable enough to formulate a fair interpretation of Japanese culture.

Unfortunately none of these issue-statements was followed up consistently or rigorously; definitely not tested empirically with objective evidence. In fact, these “working hypotheses” seemed to have been abandoned altogether right after they were so stated. In their place, he began putting forward the onslaught of harangues backed by exiguous evidence. He accuses Benedict of being:

• an “Orientalist” a la Edward Said
• a servant and an articulate spokesperson of the US occupational policy over Japan and a political educator
• a “blunt instrument” social analyst
• a writer of “political literature”
• a biased, ethnocentric anthropologist typical of her time
We should accept these claims, however sweeping, provided he could furnish us with evidence obtained through logical and/or empirical testing. Unfortunately, his evidence is too fragile to justify his sweeping overstatements; they are at very best his conjectures. It would be fair to say that there is a generic difference between a conjecture and an evidence-based statement. The former should not replace the latter, particularly in an essay intending to be analytical.

His conjectures, furthermore, appear to be driven by his particular ideological biases. An ideological motive per se should not be faulted, especially in formulating a research issue. After all, one has to start somewhere. Yet, an ideologically inspired perspective and its derivative hypothesis must be subjected to the same rigorous logical and empirical testing as any other hypothesis.

While alleging that Benedict belongs to those whom Edward Said would call “Orientalists”, Lummis seemed to have revealed his own value orientation. He noted that “flaws in the book [The Chrysanthemum and the Sword] have been difficult for many Western scholars to see,” (Italics added). A simple, yet legitimate, question needed to be raised here: what indeed motivated him to attach such a caveat? Is this just a (not-so) innocent remark to cajole or flatter the Japanese audience or does it indicate something deeper? If to be the latter, Lummis would feel quite comfortable with the school of thought which holds that the Japanese snow is so unique that only a pair of skis made in Japan, not foreign-made ones, should be used on the snow hills of Japan.

What I am alluding to here is that Lummis, consciously or not, might have subscribed to the ideological stance Ian Buruma and Avishai Margalit called “Occidentalism”, the ideological opposite of “Orientalism”. If so, is it not a foregone conclusion that “Occidentalist” Lummis found Benedict’s work less than satisfactory as he considers her to be an “Orientalist”?

His ideological biases seem particularly pronounced in two claims: Benedict was an able advocate and a framer of the US national interest in the US occupation policy over Japan. As well, she was a Western-centered, if not racist, anthropologist typical of her time.

Once again we should be reminded that it is important to consider any hypothesis with open-mindedness. This principle must be extended to those devoid of any intrinsic plausibility. However, this open-mindedness proposition must be accompanied with one indispensable condition: any hypothesis must be verified through an objective testing process in order to establish its validity. We, and particularly Benedict as she can no longer defend herself, deserve this disciplined rule of analytical discourse. At least, as Sgt. Friday of ‘Dragnet’ often said, "All we want are the facts, ma'am"

The best “evidence” he could provide seems nothing more than her association with a certain government organization, the Office of War Information, and an academic profession, anthropology. Unless he can document the specific and direct linkages sufficient to establish and verify his accusations, this is nothing more than ‘guilt by association’ or so-called “profiling”. True, more than a few early ethnographers and academics were unaware of and/or unwilling to admit their own ethnocentric or racist biases, as Lummis pointed out. However, it is more than unfair and logically incorrect to accuse her of ethnocentrisms or worse, simply because she happened to work in the milieu of the early anthropologists. Did she not write a book on race and racism, after all? Or is it sufficient for us to label someone automatically as a subservient tool of the state simply because the person happened to be a staff member of the Office of War Information? I think not. He must present evidence, not conjecture, to establish a reasonable linkage to connect her to this grandiose accusation.

Her association with the Office of War Information (OWI) seems to have heavily colored his view of her work. The OWI and the Office of Strategic Services (OSS) attracted many distinguished scholars from diverse disciplines during the World War II. That might have been in part because there were not many debates, rightly or wrongly, on the moral and legal legitimacy of the objectives of the war fought by the so-called “Greatest Generation”, definitely not to the same extent as those on the Vietnam and Iraq wars. The ideological profile of those who worked in the organizations, as expected in any human group, was not homogeneous. Granted that those think-tank organizations, in particular OSS, are considered the predecessor of today’s CIA, but they, particularly OWI, must not be seen through the lens of recent experiences of the intelligence gathering organizations of a “slam dunk” CIA nature, which has been said to have rather willingly manipulated the information to shore up the political agenda of the powers that be.

Lummis further observes that Benedict’s book “unsurprisingly for the time” “explains and justifies the defeat and occupation” of Japan. With this and other accusations, Lummis seems to have placed Benedict in the same category of the academicians who were used by or willingly served state power such as Karl Haushofer, a geopolitical theorist, whose concept was used to justify the Nazi policy of territorial expansion. My first reading of Benedict many years ago as a college freshman left me a quite contrasting impression from Lummis’s. I remember being amazed at Benedict’s perceptiveness and sensibility---characteristics of being a good poet? ---which enabled her to elicit such an intriguing, often persuasive, behavioral model of Japanese culture. I might have been naïve and probably I was, but The Chrysanthemum and the Sword, did not impress me as politically inspired, i.e., serving the US national interest, a blueprint for the “conquering power, the U.S.” to transform the Japanese nation. This seems even less plausible than claiming that Robespierre and his Reign of Terror were merely the brain child of Jean-Jacque Rousseau.

Another bias of his, a gender-based one, is worth mentioning. Lummis leaves a distinct impression that Benedict’s gender seems to have had something to do with flaws in her work. He implies that Benedict was a dreamy poet kind of a woman, (thus?) with a tendency to replace hard reality by an imagined, make-believe world. It is incomprehensible, without further evidence, why being able to write a poem, regardless of gender, should be a hindrance to being a behavioral scientist. The nature of behavioral investigation, particularly in anthropology, seems to demand the sensitivity and the perceptiveness to detect and elicit reasons why people behave as they do. She might have done the work as productively as she did because of, rather than in spite of, her poet’s ability and sensitivity to observe. Remember, she worked without the benefit of actually living in Japan for an extended period of time: she “simply” observed and collected data from mainly the Japanese-Americans in the concentration camps under the President Roosevelt’s Executive Order 9066, which was an incredible feat by any standard.

He furthermore accuses her lack of necessary methodological training, or retraining from having been an English major originally, as she spent “only” three semesters to earn her Ph.D. Three semesters may not be enough in Lummis’s mind to achieve sophistication in any matter. But some people do and Benedict seemed to have done it. I wonder if he would have made this sort of comment should Benedict happen to be a male. Would he have expressed the skepticism about the work of Malinowski, the so-called father of modern anthropology, as his training was done in natural science, rather than social-behavioral science? It is more than ironical, to put it mildly, to see someone whose basic methodology is suspect accusing Benedict’s lack of methodological training and rigor.

I am rather relieved, nonetheless, that he didn’t resort to Benedict’s slight hearing impairment as another source of her “faulty” observations. He must have known of this impairment as he rather exclusively relied on Margaret Mead’s book, An Anthropologist at Work. (Incidentally, this apparent, exclusive reliance on one source, i.e., Mead’s book, is puzzling as he claimed that he had had extensive access to the Benedict Collection at Vassar College). Mead noted Benedict’s slight hearing problem in the book, though not as weakness, but as strength as it helped her to cut off the “noises” to see the essential nature of the phenomena she was studying. I was equally, if not more, relieved that he did not attribute Benedict’s so-called faults to her Sapphic inclination, of which she was suspected. However, he used (or did not object to the editor using) the cover photos of Benedict and Mead from the controversial book by Lois W. Banner, Intertwined Lives : Margaret Mead, Ruth Benedict, and Their Circle, which was sensationalized when it came out on account of its treatment of the two women’s relationship, which was said to have been more than just good friends.

II. Epistemology and other discrepancies

Even if Lummis could have come up with objective evidence to buttress his claims, there still remains a fundamental, epistemological gap which seems to impede making a balanced assessment of Benedict. Lummis’s perspective seems to have derived from a conceptual base fundamentally different from Benedict’s. This naturally contributed to accentuating the differences, rather than finding the congruencies, between them. To begin with, in the absence of his own definition, even a provisional one, we are never quite sure in what sense Lummis is using the term “culture”. However, the tenor of his argument, including his choice of the title of the essay, hints that culture in his mind is an entity which is static, concrete and absolute, while Benedict uses it from the perspective of behavioral, dynamic relationships. Another, even more significant, epistemological discrepancy may be in the understanding of the notion of the “ideal model” in the social and behavioral science research.

An ideal model in social science research is not, nor does it intend to be, an exact replica of a given social-cultural reality. Rather it is an analytical schema depicting the elements deemed essential and their interrelationships in defining the essence of a given phenomenon. It is a heuristic tool for observation and analysis, leading to the formulation of (testable) hypotheses. Through the verification of those hypotheses, we try to obtain a new insight into the particular phenomenon we investigate. An effective analysis was produced through the use of an ideal model by such scholars as Max Weber on the role of the Protestant ethic in the rise of capitalism, Robert Bellah in his analysis of the so-called achievement orientation in Tokugawa Japan as the key value in Japanese modernization or, for that matter, Benedict’s shame culture in analyzing Japan.

The lasting contribution of Benedict, therefore, should not be determined solely by descriptive “accuracy”. Trying to find exceptions in the descriptive narratives of Benedict seems to be misdirected, though the detractors of Benedict seemed to exercise this option often. The importance of her work is its ability to guide how to observe, conceptualize and thus to analyze the dynamics of Japanese culture in a more objective and systematic manner. Once a pattern of a given culture is defined in such a manner, we could expand our horizon beyond simply developing a rich case study and engage in comparing cultures systematically and objectively. She left us with an analytical perspective and a research tool enabling generations of scholars of Japan and Japanese culture to collect objective and comparable data to bring forth new insight. Unfortunately this aspect of Benedict’s legacy never entered into Lummis’s critical thinking. When he grudgingly concedes the usefulness of Benedict’s work, he does so only “as a work of political literature”, but never as a heuristic analytical model.

It is also important to remind us of the sociological impact of Benedict’s work, which might be even more significant than her epistemological contributions. She demonstrated that we can look at Japanese culture, or for that matter any other foreign culture, as it is, through grounded observations, rather than simply adopting the emotive images dictated by the prevailing notions of Japan and her culture. She strived to achieve objectivity by marshalling whatever empirical data was available to her within the wartime constraint. Though we take it for granted nowadays, she did all of this in the middle of what John W. Dower called a “War without Mercy”, in which racially charged emotions against the Japanese were intense, intense enough to persuade the American people to acquiesce in the implementation of Executive Order 9066, the internment of the American citizens who happened to share Japanese ancestry.

Her work seems to have proved that the culture of another country, even that of the enemy nation in the midst of war, could and should be analyzed objectively. She demonstrated the possibility of observing Japan and her culture without peering through the silk screen of Mount Fuji, geishas and cherry blossoms, or worse, the war time caricatures of the Japanese, all of which frequently misled people to see Japan in unrealistic terms, both positively and negatively. We could even venture to say that Benedict seemed to have achieved the paradigm change on how to observe Japan and other non-American cultures. Benedict could have taken an easy way out by appealing to the then prevailing image of the Japanese and their culture, particularly if she was so eager to produce a readily usable work of “political literature“, as Lummis noted.

III. Opportunities Missed?

Lummis’s locating and subsequent interviewing of Robert Hashima, allegedly Benedict’s exclusive confidant-informant, could have allowed him to make a unique contribution to the study of Benedict and her legacy.

Mr. Hashima, according to Lummis, was the authoritative sounding board for Benedict on Japanese culture. Lummis was able to interview him twice over a two year period from 1996 to 1997, which led him to conclude that Benedict placed her trust in and reliance on the wrong person. To him, the informant is ethnically Japanese, but a somewhat alienated transplant to Japanese society, hardly suitable qualifications for being the exclusive and authoritative window on Japanese culture for Benedict or anyone else. As a result, Benedict’s work, which is said to be heavily relied on the informant, is fundamentally unreliable.

To prove this conclusion requires testing two logically connected propositions: first, Mr. Hashima was indeed the exclusive and authoritative source of information for Benedict and heavily influenced her thinking on Japan and its culture, and second, the qualifications and background of the informant were unsuitable for providing the decisive voice on Japanese culture. He attempted to do so through a face to face interview with the informant.

So far as we can tell from the interview results presented in the essay, the informant appears a modest, sensitive man, hardly one to glorify his role in the research project under Dr. Benedict, as he called her. Moreover, the interview results seem to suggest that the informant seemed rather uncomfortable in dealing with Lummis’s direct questioning about his role as the exclusive source of information for Dr. Benedict.

In part based on my own experience in survey research and in-depth interviewing in Japan, there are ways, and needs, to manage the delicate situation in ascertaining sensitive, personal information from a rather reluctant interviewee. Though no apparent linguistic barrier existed, Lummis could have engaged, for instance, a Japanese interviewer to conduct the interview of the seemingly less forthcoming Japanese person. In general, such an approach helps ease the apprehensiveness of the interviewee and makes it easier to elicit personal, guarded information. In addition, instead of asking embarrassing questions directly as Lummis seemed to have done, he could have resorted to gathering indirect, yet objectively verifiable, information. He could have asked how or through what processes the informant provided input to Benedict; how often and intensive his interactions with her were in the research project, and who else, if anyone, was involved in supporting Benedict’s work, and what roles did other people play, etc. Indirect questioning would have allowed Lummis to triangulate the answer to his ultimate question regarding the informant’s role in shaping Benedict’s understanding in a more objective way. As it was, the interview results reported in the essay do not prove conclusively Mr. Hashima’s exclusive role in shaping Benedict’s thinking.

There is a prima facie plausibility in Lummis’s doubt about the qualification and the authoritativeness of the informant on Japanese culture, the reasons being his late introduction to Japanese culture and his assumed general alienation from it. However, is it not equally plausible that someone who came in contact with the foreign (in this case, Japanese) milieu later might have a heightened sense of cultural difference? The informant might have had sharper understanding and awareness of Japanese culture precisely because of his late and interrupted introduction to Japan. Unless you are taking an “Occidentalist” stance of some sort (as Lummis seems to do, i.e., only “real’ Japanese can understand Japan?), it is not certain that Mr. Hashima was proven ill- qualified to understand Japanese culture.

In short, from the confines of the interview results presented in the essay, Lummis proved neither of the two propositions: they are at best inconclusive. We never know whether Mr. Hashima was really the exclusive confidant who significantly shaped Benedict’s thought process. It would be misleading to make any further inference that Benedict’s work is flawed because she engaged Mr. Hashima as the exclusive and authoritative informant on Japanese culture.

I happen to believe strongly that Lummis had a golden opportunity to prove his points through his encounter with the so-called Benedict’s confidant and provide new interpretations of Benedict and her work. Yet the lack of analytical rigor in his approach seemed to let the opportunity to slide by.

IV. Concluding Remarks

One of the memorable episodes in Robert M. Pirsig’s Zen and the Art of Motorcycle Maintenance involves a light, yet intriguing debate at a social evening. It is about the seemingly odd first line in the instruction manual for assembling a Japanese bicycle. It said that the very first step in assembling a motorcycle is to be in a calm and peaceful state of mind. The practical, rationalist camp of the evening thought it amusing, if not silly, as it does not directly touch on the specific steps involved in the assembly process itself. The other camp took it otherwise: it is appropriate and even necessary to have the words on the mental preparation right at the beginning of the manual. Without the proper frame of mind, it would be much more difficult to tackle the detailed tasks and deal with inevitable complications to follow at the various stages in assembling the motorcycle.

An analogy can be made as to one’s mental preparation in carrying out a critical assessment of a book. One must start with fairness and open mindedness. Any preconceived judgment should be left behind by the time the assessment commences. Unfortunately this simple, seemingly reasonable instruction would not have impressed Lummis. Instead, he began, as noted in his preface, with the determination to find fault with Benedict’s work.

Though having been impressed initially with Benedict’s work, he somehow reached the conclusion that, having a decent relationship with the Japanese people, would be difficult “unless I could drive this book and its politely arrogant world view, out of my head”. This motivation is admirable for someone trying to immerse himself in a new cultural milieu. T.E. Lawrence (of Arabia) provided a similar piece of advice to his fellow British officers coming to the Middle East: leave their Englishness in England. However, this sort of determination seems inappropriate as it is counterproductive in conducting a critical assessment of the intellectual contributions of others. His single minded pursuit of faults in Benedict’s work inevitably contributed to his tendency to reach sweeping and prejudicial judgments without benefit of evidence. This was unfortunate particularly when we know he had potential grounds to make a fair and even unique analysis of Benedict’s legacy. Had he not squandered those golden opportunities or at least had he begun his analysis of Benedict without preconceived judgment, all of us would have been enlightened further about Benedict’s contributions to the understanding of Japanese culture and Japan itself.

Toru Uno, Ph.D. in political science, is a former executive in the Canadian Federal Government. Prior to joining the Canadian public service, he taught political behavior, Japanese politics, and analytical methods at San Francisco State University and the University of Tennessee as an assistant professor. He also taught political science in a Canadian university.

How to Critique: Indeed

C. Douglas Lummis

I am very sorry to learn that my essay on Ruth Benedict’s The Chrysanthemum and the Sword has brought Mr. Toru Uno such emotional distress, leaving him in a state of “disappointment bordering on futility.” But I am glad that he has recovered sufficiently to put the reasons for his distress into words and to engage me in a debate, as this presumably means that he thinks discussion might not be futile after all. I certainly hope he is right about that.

I have read his letter several times, and, following his good suggestion, I will try to answer his criticisms in a calm state of mind.

This will take some doing, as Uno’s letter itself does not strike me as presenting a model of tranquility. And, yes, I can give evidence to support this. Consider the language he has used to describe my essay. He says it is “sweeping,” “overstated,” “speculative,” “ideological,” “erroneous,” “skewed,” “conjectural,” “disjointed,” “grandiose,” a “harangue,” “biased,” “occidentalist,” possibly sexist, based on “exiguous” evidence, and uses the methods of “guilt by association” and “profiling.” He says that it fails to follow the fundamental rules of social science research which any college freshman should know, and he is kind enough to explain what those rules are, namely to begin with one or more testable hypotheses, and then to support these with factual evidence.  In this essay, Uno says, there are “no facts”.

Now that’s quite a lot of things to be wrong with an essay. And I must say, it’s rather disrespectful to lecture a retired professor of political science on the basics of research. But I’m afraid I must also begin (Stay calm, Lummis! Stay calm!) with a brief lecture on methodology. You see, Mr. Uno, this essay is not the report of a behavioral study of Japanese society. Its purpose is not, say, to determine whether indeed Japan is a “shame culture.” It is an analysis of a text. And in textual analysis, the factual evidence is the text. Once you understand that, you will see that the essay is loaded with “facts”. At the end there are 38 footnotes, most of them to textual quotations. These quotations are facts: it is a fact that they were written as I quoted them. And these facts do logically support the hypotheses on which the essay was based. (Of course, they don’t “prove” it; there is no such thing as absolute proof in these matters.)

What, then, are the hypotheses, and what is the evidence? Uno has attempted a list, but it is inaccurate and incomplete. (He wrote that “it is no easy task to follow Lummis’s argument”, but I fear the case may be worse still, namely that he has not succeeded in following it at all.)

The main hypotheses fall into two categories. The first are specific critiques of Benedict’s Chrysanthemum and the Sword.
1) Benedict confuses ideology with culture.
2) She treats Japan as though it were not a class society. Thus, she takes the modernized version of the bushi ethic as representing Japan as a whole.
3) She sees Japanese culture as essentially static. More precisely, she sees it as incapable of change from within, though it can be changed by intervention from outside (the Occupation).
4) She claims that Japanese ethics are entirely based on shame, and that guilt plays virtually no role there
5) She claims that American culture is the almost perfectly matched opposite to this.
6) As the combined consequence of 1~5, the book is not merely ethnocentric, but gives ethnocentrism a new basis, replacing race with culture.

The second set of hypotheses has to do with the factors that may have caused her to come to these conclusions, i.e. problems of methodology.
7) She used an anthropological method that was developed for the study of small-scale societies to investigate a large-scale industrial society.
8) Her attitude toward Japan, and toward anthropology as a whole, was influenced by patterns of thought revealed in her poetry and in related notebook entries.
9) She was strongly influenced by the views of Robert Hashima.
10) Because of the peculiarities of his experience in Japan, Hashima was persuaded that the highly schematic and ideological picture of Japanese culture taught in the schools during the militaristic period, was “Japanese culture” itself. And Hashima in turn persuaded Benedict that this was so.

I could probably ferret out from the text a few more secondary hypotheses, corollaries, etc., but this should do for now. Now let’s look at the evidence, following the same order.

1) (Culture and ideology) Here Uno scores maybe half a point, as for evidence I have relied largely not on direct evidence but on authority, by citing three giants in post-war Japanese scholarship (Tsurumi Kazuko, Watsuji Tetsuro, and Yanagita Kunio) who were among the first to point this out. But further direct evidence (of which the above three were not aware) is provided by the fact that Robert Hashima’s model of Japanese culture, which he himself says he learned from the Education-Ministry-controlled curriculum in high school and teacher’s college during the militaristic period, found its way directly into the pages of Benedict’s work.

2) (No classes) This is clear from Benedict’s statement that the ideal informant for her research could be “anybody”: a titled aristocrat, a member of the Privy Council, a factory owner, a teacher, an army general, an assembly line worker, or a farmer (not to mention, a woman) could all be expected to give about the same answers. You would not say this if you knew you were dealing with a class society. But the main evidence is the entire text of The Chrysanthemum and the Sword: no analysis of class can be found within it.

3) (Static culture) If Benedict didn’t see Japanese culture as essentially static, she wouldn’t have been able to use examples from the Edo period to explain behavior in the mid-20th century. More to the point, she argued directly that as a shame culture, having no real principles, Japan is incapable of self-criticism and therefore of change from within. But she did believe that Japan could be changed from outside (albeit only if this is done with some delicacy), in particular by the Occupation whose task, she says, is to “break up” the harmful patterns of Japanese culture.

4) (Shame culture) It is obvious to anyone who read the book that Benedict made this assertion. I did not see it as my task in this essay to prove that Japan is not simply a shame culture. I did, however mention Yanagita Kunio’s observation that in order to come to this conclusion, Benedict had to avoid mentioning Japan’s elaborate vocabulary for expressing guilt – or, more likely, was unaware of it.

5) (Guilt culture) She said this directly in the lines that I quoted. But it is true that, other than me and Clifford Geertz, not many readers have noticed that she sneaked into the book a highly dubious model of American culture without offering a shred of evidence in support of it.

6) (Cultural ethnocentrism) Uno accuses me of calling Benedict a racist, though I specifically said she was not. I did say she was ethnocentric. The evidence for this is the combined force of her arguments. She presents America as the model of a society founded on freedom and principle (guilt culture), i.e as Japan’s binary opposite. And in her metaphor of the bonsai returned to nature, she implied that for a Japanese to adopt American culture would bring all gain and no loss: pure, natural growth. And though she does not say so, it is built into the structure of her metaphor that for an American to adopt Japanese culture would be like stuffing a full-grown tree into a bonsai pot: grotesque disfigurement and probably impossible. Is it wild speculation to call that ethnocentric?

7) (The anthropological method) This is a restatement, from a slightly different perspective, of hypothesis 2). I’m not sure that it needs to be proved; rather I think the burden of proof would be on someone who doubts it. Nowhere does she say she has changed her method. Treating a society as though it were homogeneous might be appropriate to a small-scale tribal culture, but certainly not to industrial Japan.

8) (Influence of Benedict’s poetry) This assertion is supported indirectly by statements by Margaret Mead, who perhaps knew Benedict’s mind more intimately than anyone else, and directly by quotations from her poetry and notebooks, supported by an argument showing the relation between these and her anthropology. That’s how content analysis works.

9) (Influence of Hashima) It’s puzzling that Uno says I gave no evidence for this. There are three major pieces of evidence: i. Hashima himself said so; ii. Hashima is the only informant thanked by Benedict in her acknowledgements; and iii. In Benedict’s notes on her interview with Hashima you can see the pattern of “shame-culture Japan”, i.e. the heart and soul of The Chrysanthemum and the Sword, taking shape. This last I think comes close to being a smoking gun, except of course that none of this is a crime.

10) (Hashima’s view of Japan) The evidence for this is my interviews with Hashima, quoted at length.

Concerning these interviews, Uno is certainly correct that they could have been better done: there are other questions I could have asked and, yes, it’s quite possible that sometimes I was sometimes too blunt. But no, I did not find Hashima to be overly shy or modest. He spoke quite respectfully of Benedict, but he had no hesitation in claiming himself to be her chief informant. He even said, in his jovial manner, that as (as he believed) the Emperor system was saved by Benedict’s intervention, and that as (as he believed) it was he who gave Benedict this idea, he should be given a medal by the Imperial Household Agency.

It might have been good, as Uno suggests, to have organized a follow-up interview with a Japanese interviewer. But I rather suspect that, given the harshness of his critique of Japanese culture, it would have been more difficult for him to speak frankly with a Japanese interviewer than with me. Be that as it may, he did say what he said, and that’s a fact.

Given that Uno’s devotion to methodology is so great that lapses in rigor discovered in other people’s works can cast him into a state of futility, it is surprising to discover in his letter so many careless and speculative accusations. Because I wrote that “the flaws in the book have been difficult for many Western scholars to see” he says I “might” be guilty of ‘Occidentalism’”. (Actually I wrote this because I believe it to be a fact.) Because I emphasized Benedict’s poetry in my analysis, Uno says I leave a “distinct impression” of gender bias. (Here he misses my point entirely.) He writes that after reading the piece, “it became somewhat easier to appreciate the dilemma his earlier editors had to deal with” (that is, the editors who revised it without my consent) apparently without looking at the revised version to see what changes they actually made. And because my article in Japan Focus was illustrated with a picture of the cover from Lois Banner’s Intertwined Lives: Margaret Mead, Ruth Benedict, and their Circle, which contains controversial material on the two women’s relationship as lovers, Uno includes a vague innuendo which I guess is meant to suggest that the inclusion of the picture is intended as a slander against Benedict. Now this is real bottom-of-the-barrel innuendo mongering. In fact the picture was selected by the editors of Japan Focus, and I saw it for the first time when the article came online. But Uno is right, I did not and do not object to it. It serves to remind readers that interest in Mead and Benedict remains very lively today, and also it shows very nice pictures of the two. If there is any slander here, it is Uno’s backhanded slander against Banner, suggesting – what? – that her book is something that ought not to be shown in public? It is some years now since gay and lesbian studies have come out of the closet and entered the space of public discourse, and I understand that Banner’s book, which I am sorry I have not yet read, is an important and scholarly contribution to that field, as well as to the field of anthropology generally.

Uno also spent a full page trying to argue that the fact that Benedict did her research for the book while working for the OWI does not count as evidence that the result is likely to favor the U.S. government’s position. This is what he calls “guilt by association” and “profiling”, and he suggests that it is equivalent to calling her a Nazi collaborator. But the issue here is not “guilt”; I have never claimed, and do not claim, that Benedict was guilty of immoral or criminal behavior. She was doing the job she was hired to do. If one is hired as a policy advisor, it is, as I wrote, unsurprising that the resulting research will be policy advice. But it means that we who are not policy advisors, and whose position in the world is very different from that of a policy advisor, need to look very carefully before swallowing such research whole.

It is also notable that the rigorous Dr. Uno gives as evidence to support his belief that Benedict’s book contains no pro-government bias, the fact that it did not seem so to him when he first read it “many years ago as a college freshman.” He has not told us whether he has given it a second reading more recently; if he has, it is curious that he would refer only to the first. But as I wrote in the introduction to my essay, when I first read the book, also many years ago, I had the same impression as Uno. I thought it was a model of liberal tolerance, brilliantly executed, and just right. I still think it is brilliant, but I no longer think it is a model of tolerance, or just right. The change came about through the experience of living in and studying the place she was writing about, and the experience of giving the book a second, third, etc., reading. So Uno’s advice that one should approach a book calmly and without presuppositions, apt in some cases, is a little hard to apply to this one. I was calm enough when I first read the book in 1960, but pretty troubled by it when I began to reanalyze it in 1970. It was too late to empty my mind of all previous notions of the book when I began to re-read it, and also I got too passionately involved in the project to remain in a state that Robert Pirsig might call calm. But then, putting together a piece like this is not much like assembling the manufactured parts of a motorcycle.

Finally, though saying this may only reveal my ignorance of the inner passions of a devoted methodologist, I find it difficult to believe that Uno’s perception that my essay was methodologically flawed could by itself have been sufficient to so upset him. Usually when we read an article that we think is based on flawed or insufficient evidence, we just skip it and go on to something else. Is it really possible that Uno’s distress comes only from the article’s methodology, and not at all from disagreement with its conclusions? Or if he does disagree with some of the conclusions (and I do believe some of the conclusions are upsetting to some people) would it be too much to ask him to tell us what these are? I think if we had that information, we could have a much more interesting discussion.

C. Douglas Lummis is the author of Radical Democracy and is a Japan Focus associate. He taught at Tsuda College.

This exchange between Uno Toru and C. Douglas Lummis was posted at Japan Focus on December 4, 2007.