Translator Lounge

Introduction

This study illuminates the role of translation in the examination of the validity of version one of the Intercultural Development Inventory (IDI) (Bennett & Hammer, 1998) and, by extension, some assumptions surrounding the cross-cultural transferability of theoretical frameworks such as the Developmental Model of Intercultural Sensitivity (DMIS) upon which the IDI is based.

The Intercultural Development Inventory (IDI) is a psychometric instrument that operationalises and measures the DMIS. Because the IDI was available only in English it was first necessary to translate it into Japanese so that it could be used with its target population. However, the translation process raised many questions about whether the concepts that comprised the instrument, and by extension the DMIS, were transferable across languages and cultures.

The translation protocol used for this study was a five-step procedure, much more rigorous in nature than the widely-used translation/back-translation protocol, commonly known as the ‘Gold Standard’ for translation. The added step of having translators use a draft translation to perform a task outside the translation context was a contribution to the art of translating and adapting instruments.

The Model

As a theoretical framework, the DMIS is Milton Bennett’s (1986) attempt to describe the process through which one increases one’s intercultural sensitivity. Developed as grounded theory, based on long years of field observation, it describes the stages through which people move in acquiring increasingly sophisticated understanding of cultural difference, beginning with three ethnocentric stages and continuing through three, more inclusive, ethnorelative stages.

The first stage, Denial, is the default end-product of socialisation into one’s native culture. In Denial, a person is insensible to the existence of other cultural norms and worldviews, understanding only his or her own culture or way of thinking to be the natural order of things. The label Denial refers to the denial of the reality of other cultural perspectives or values, on any level worth considering.

When it is no longer possible to ignore the existence of other worldviews or cultural points of view, the natural reaction is to first treat them with hostility and suspicion. This is the stage that Bennett labels Defence. Defence can take the form of actively denigrating other cultures or merely asserting the superiority of one’s own. It is the worldview that underlies exasperated expatriate statements that begin with, “Why can’t these people learn how to…?” or “Why do they do everything backwards here?”

Further exposure to and observation of other cultural points of view can lead to the third ethnocentric stage, Minimization. In Minimization, one views differences in clothing, food, language, or customs as superficial, beneath which people are really pretty much the same. This sounds like a laudable, egalitarian point of view until one remembers that from an ethnocentric worldview, the standard for the sameness of people can be only oneself. This worldview still does not allow for other cultures to have an independent reality or validity.

Moving beyond the ethnocentric stages, a person progresses to the DMIS stage of Acceptance, which, as the label implies, brings the realisation that other cultures do have an external reality and are valid on their own terms.

Acceptance lays the foundation for Adaptation, first Cognitive and then Behavioural, in which other cultures’ frames of reference and modes of acting are incorporated into one’s own cognitive and behavioural repertoire, permitting one to act as a true bridge between and interpreter of other cultures.

Figure 1 Continuum of Intercultural Sensitivity

To explain the mechanism by which a person progresses through the stages towards increased intercultural sensitivity, the DMIS is a social-constructivist model that posits that it takes more than exposure, even of a prolonged nature, to alter one’s worldview. As George Kelly (1963, oft cited by Bennett in workshops and lectures) observed:

A person can be witness to a tremendous parade of episodes and yet, if he fails to keep making something out of them…he gains little in the way of experience from having been around when they happened. It is not what happens around him that makes a man (sic) experienced; it is the successive construing and reconstruing of what happens, as it happens, that enriches the experience of his life (p. 73).

In other words, it is not enough to be in the vicinity of events as they occur, as Bennett likes to remind us. One must interpret and make sense of those events in order to learn from them, through a process of reflection. In the context of intercultural sensitivity, events must be observed in order to yield data about other cultures, but those data must then be interpreted (reflected upon) to yield information and then knowledge. Conscious reflection upon the feelings and reactions the experience engenders, both towards other cultures and towards one’s own, contributes to the growth and evolution of one’s worldview.

The social constructivist nature of the model has particular resonance for translators as the nature of social construction varies from culture to culture, and by implication, from language to language.

The question of whether the DMIS and the IDI are transferable and applicable across cultures and across languages is an empirical one. To illustrate, in validating a measure of cross-cultural adjustment, the Intercultural Adjustment Potential Scale (ICAPS), Matsumoto et al. (2003) acknowledged that the transferability of their instrument and the underlying psychological traits was subject to empirical verification. Although they argued that:

…the same psychological skills may (italics added) be necessary for intercultural adjustment of any individuals from any culture as they adjust to a different culture because the psychological skills underlying the process of managing intercultural stress and conflict may (italics added) be the same regardless of culture even though the manifest content of the conflict is culturally specific” (p. 545)

they acknowledge that, “We do not know (italics added) whether or not the ICAPS can predict intercultural adjustment for individuals other than Japanese” (p. 545), and went on to report six studies that investigated the question. Issue could be taken with their casual attitude to the translation of the English original into Japanese, “(A)ll measures were translated into Japanese and their accuracy was verified using back-translation with no problems” (p. 546) and a rationale for a more comprehensive translation protocol is proposed later in this paper. However, they acknowledge, by translating the instrument that language has the potential to significantly affect the results obtained.

Linguistic Determinism
In 1929, Edward Sapir first advanced the ideas that he and his student Benjamin Whorf would later propose as the theory of linguistic determinism. In a classic passage, Sapir (1929) argued that:

Human beings do not live in the objective world alone, nor alone in the world of social activity as ordinarily understood, but are very much at the mercy of the particular language which has become the medium of expression for their society. It is quite an illusion to imagine that one adjusts to reality essentially without the use of language and that language is merely an incidental means of solving specific problems of communication or reflection. The fact of the matter is that the ‘real world’ is to a large extent unconsciously built upon the language habits of the group. No two languages are ever sufficiently similar to be considered as representing the same social reality. The worlds in which different societies live are distinct worlds, not merely the same world with different labels attached. . . We see and hear and otherwise experience very largely as we do because the language habits of our community predispose certain choices of interpretation (p. 69).

Sapir and Whorf were ahead of their time in understanding that reality is constructed and that one must have categories for classifying experience, otherwise it (the experience) is meaningless. As Whorf (1956) put it:

We dissect nature along lines laid down by our native languages. The categories and types that we isolate from the world of phenomena we do not find there because they stare every observer in the face; on the contrary, the world is presented in a kaleidoscopic flux of impressions which has to be organised by our minds. . . We cut nature up, organise it into concepts, and ascribe significances as we do (pp. 213-214).

The Sapir-Whorf Hypothesis states, in its strong form, that language determines thought; that language and thought are identical. Linguistic determinism attracts few followers today since there is strong evidence against it, including the possibility of translation between languages. However, the weak version of the Sapir-Whorf Hypothesis, that language strongly influences thought by shaping our categories for classifying experience, is much more widely accepted.

In order to pursue a research agenda based on the DMIS as a theoretical framework, and the IDI as an empirical instrument, it was first necessary, therefore, to examine the reliability of the IDI in Japanese translation, and the validity of inferences from data gathered from a population that is culturally and linguistically very different from the population on which the original statistical analyses were performed.

Procedures for Translating the IDI
Bennett and Hammer (1998), in reporting their IDI development and validation procedures, stressed the cross-cultural applicability of the instrument (i.e., that a sufficiently diverse cross-section of different languages and cultures had been represented in their subject pool to eliminate any possible effects on the results from cultural and native-language differences) so it became important to test that assertion by privileging a translation that was faithful to the wording and concepts of the original. However, a number of the translators working on the project commented on the ‘foreignness’ to the Japanese mind of some of the concepts appearing in the instrument. These items were flagged for attention in the statistical analyses.

There is also an important difference between translating and adapting instruments. Ideally, the process of transferring an instrument from one language to another should be one of adaptation, in which the pith of the theoretical concepts is given precedence. As Hambleton and Patsula (1998) explain:

Test adaptation includes such activities as (1) deciding whether or not a test can measure the same construct in a different language or culture, (2) selecting translators, (3) deciding on appropriate accommodations to be made in preparing a test for use in a second language, and (4) adapting the test and checking its equivalence in the adapted form (p. 155).

Test adaptation also sometimes requires generating a new item where conceptual equivalence is difficult to achieve (Kristjansson, Desrochers, & Zumbo, 2003).

Although the process followed incorporated many of the points enumerated by Hambleton and Patsula (1988) above, the Japanese-version IDI created for this study was the product of a conscious process of translation rather than adaptation. For example, judgments on conceptual equivalence and accommodations were consciously avoided because the cross-cultural transferability of the original instrument and its concepts was the object of empirical verification in the study.

The IDI was first translated into Japanese for an earlier study (Greenholtz, unpublished) that attempted to determine the level of English proficiency, as measured by the Test of English as a Foreign Language (TOEFL), necessary to meaningfully complete the IDI in English. It was the lack of any clear correlation between TOEFL scores and evidence of ability to complete the IDI that cemented the decision to use a Japanese translation for this study.

Briefly, that original study asked 100 native-Japanese-speaking university undergraduates at various levels of English proficiency each to translate ten of the sixty IDI items into Japanese. They were also asked to state their level of confidence with their translation, in an effort to indirectly identify concepts in the instrument that might pose extra-linguistic (conceptual) difficulties for Japanese speakers. The translations were rated and statistically analysed with TOEFL scores as the independent variable. As noted earlier, no clear correlation with TOEFL scores emerged with predictive value from those data.

In order to rate the undergraduates’ responses for that study, it was necessary to create a Japanese IDI ‘master’ translation.

Four Japanese-English bilinguals (native Japanese speakers) were each asked to independently translate all sixty items into Japanese. They were judged to be bilingual because they had each completed post-graduate work in the United States (one at the doctoral level). One of the translators, who had completed an MBA at an American university, was working in the United States for a large Japanese multinational corporation. Two others, one of whom was also a certified IDI administrator, were experienced English teachers working in Japan. The fourth translator had completed a PhD from Columbia University in second-language acquisition, and was also living in the United States.

This researcher’s qualifications for leading the translation team were as follows: native English speaker who has spent over half of my adult life in Japan; passed the highest level of the Japanese Language Proficiency Test (Nihongo Noryoku Shiken, Japan Foundation) in 1984; have done a large number of Japanese-to-English translations and published an English-to-Japanese translation of My Friend David (Edwards and Dawson, 1983) entitled Mai Frendo Deibido (Greenholtz and Morita, 1988); certified as a Japanese-English court interpreter by the government of Ontario in 1991; and serve as a consecutive interpreter in a variety of contexts. I am also a certified administrator of the IDI.

A final version was selected from the pool of items (the four independent translations) and this version was given to two raters, native-Japanese-speaking doctoral candidates at a Canadian university, to rate the undergraduates’ translations. The rating exercise required the raters to consider the items much more deeply than simply comparing them to the English originals or performing a back translation would have, because they had to justify the score they had assigned to each of the translations.

Following the rating phase, each item was discussed with the two raters for fidelity to the language and concepts of the original, within the constraints of natural Japanese. Several changes to wording of items were suggested and many of those changes were incorporated into what became the Japanese-language master translation used in this study.

This procedure far exceeded the ‘translate/back-translate’ procedure that had long been considered the ‘Gold Standard’ for translating material into other languages. The method employed exceeded even the ‘New Gold Standard’ described in Kristjansson, Desrochers, and Zumbo (2003) who (citing Behling & Law, 2000; Hambleton & Patsula, 1998) point out, “direct translation and back translation can deal with literal meaning only”, and “(B)ack translation cannot detect differences in conceptual understanding of the question, and so cannot ensure psychological equivalence of the items in a scale or questionnaire” (p. 135). Although the intention of the study was to use as literal a translation as possible, it also needed to be cognizant of conceptual inconsistencies or difficulties posed to Japanese speakers by the IDI because the final goal of the study was not only to produce a translation. It was also to examine whether there was a sufficiently strong evidentiary validity argument for using the translation. As part of that process, the conceptual nuances helped to make sense of the statistical analysis.

For example, discussion with Japanese-English bilinguals has revealed that it is difficult for them to provide clearly distinct lexical alternatives (‘right’ and ‘wrong’ usually comes out first as some variation of ‘correct’ and ‘incorrect’, i.e., as adjectives rather than nouns). This is not to say that lexical equivalents cannot be found, only that they must be actively searched for, or perhaps more context provided in the sentence. One translator suggested that without a complement the concept would prove significant syntactic difficulties in Japanese.

As an additional example, one of the translators, commenting on a particular item and concepts differentially available in English and Japanese said, “Another mutsukashii (difficult) sentence, referring to culture in terms of superior and inferior, although it makes sense in English” (personal communication).

The translation process is summarised in Table 1, below.

To augment and further verify the translation, a research assistant (RA), a Japanese-English bilingual undergraduate student, completed both the English and Japanese versions at one sitting. Her scores were compared on an item-by-item basis and where they were different, an attempt was made to determine why. Some of the differences were due to items that had concepts that did not transfer well to a Japanese cultural framework (as flagged by the translators, see discussion below), but were conceptually available to her in the original English version.

Given the value of the insights produced, this technique (having a bilingual complete both versions of the instruments at the same sitting and comparing the results on an item-by-item basis) should be added to the protocol for translating/adapting instruments.

Methodology

Four hundred students completed the Japanese version of the IDI, over a three-year period. The students were all undergraduates at a large private university in Japan. Two hundred eighty-seven (287) were female and 113 were male. Three hundred and seventy-one (371) of the students were in second year, twenty-seven (27) in third year, and two (2) were fourth-year students. Participants ranged in age from nineteen (19) to forty-eight (48) years. Ninety-five (95) percent of the participants were between nineteen and twenty-one years of age.

These IDIs provided the data for assessing the construct validity of the Japanese version.

The 400 IDIs were analysed using principal components analysis, first with Direct Quartimin rotation (see Zumbo & Taylor, 1993) to determine whether there were any correlations among the factors. Once it was confirmed that the inter-factor correlations were small, Varimax rotation was used to determine whether the factor structure of the Japanese version matched that of the English original, for this population. The results were further confirmed with a factor analysis (Maximum Likelihood with Varimax rotation).
A hybrid exploratory/confirmatory approach was used for the principal components analysis. It was not entirely exploratory because the results of the analyses of the original English IDI guided the determination of the general parameters for this analysis. However, since it wasn’t a certainty that the Japanese version would be identical to the original, the analysis was not wholly confirmatory either. In addition, following Kelloway’s (1998) guideline that confirmatory factor analysis requires a 10:1 subject to item ratio, the data set, with only 400 subjects for a sixty-item instrument was not large enough for a purely confirmatory factor analysis.
Principal components analyses were performed for four- through seven-factor models to find the best fit. Knowing that there ought to be in the neighbourhood of six factors permitted the selection of initial Eigenvalues greater than 1.5 (rather than the more traditional rule-of-thumb of 1.0), which greatly simplified the analysis.
The four-, five-, six-, and seven-factor models were analysed and interpreted using the rotated factor matrix, the scree plot, and nonredundant residuals with absolute values greater than 0.05, to determine which made the most sense both statistically and interpretively.

Results

Results from the DMIS stage of Minimization will be adduced to illustrate how language and culture issues were illuminated by the translation process, without going into all of the details of the statistical analyses that might not be of interest to a readership comprised of the translation community. One of the subscales of Minimization is called Transcendent Universalism, which taps the belief that humanity is united by a Creator, whatever its surface cultural manifestations. However, all of the Japanese informants in the study (the translators, the RA, and the raters who helped to refine the items), unanimously noted that the Transcendent Universalism items were conceptually nonsensical in Japanese culture, despite the fact that the words could be translated. As one of the translators put it in private e-mail correspondence “I don’t understand this whole sentence … M [another translator] avoided translating all ‘the children of a spiritual being’ [items]. I don’t know what that is either. It’s possible to put Japanese words together as a translation without understanding what that means. demo hen” [but it’s weird – author’s translation].

Yamamoto and Tanno (2002) cited a similar difficulty. They said, (in translation from the original Japanese by this author) that:

I made an effort to be as faithful to the original wording of the IDI as possible, but some expressions came out as difficult to understand in translation. However, it wasn’t a simple translation problem. It is possible that some of the items were difficult to understand culturally. For example, in the Universal Values subscale of the Minimization stage items 12, 23 and 59 are phrased as ‘At our root is a supernatural holy being (choshizende shinseina sonzai), but the English original for ‘supernatural holy being’ was ‘spiritual being’. I thought about translating it [in a syllabic katakana rendition] as ‘supirichuaru bi’ingu’, but trying to be faithful to the original, I checked with Bennett and as a result I used ‘supernatural holy being’. However, some doubt remains as to whether the value held by those who believe we are united under a supernatural being is an appropriate expression of what the Minimization stage of intercultural sensitivity means to Japanese (p. 40).

Although the most glaring examples were the items that comprised the Transcendent Universalism subscale of Minimization, clearly products of a Judeo-Christian mainstream culture, this was not a quibble over three items out of sixty. It was fundamental to the way the instrument was constructed; out of quotes from actual interviews, not fabricated items.

Translation also played a critical role in bringing other issues came to light. One recurring issue was the IDI’s references to ‘being a member’ of one’s own culture as juxtaposed with people from other cultures. Although some of the phrases sound a bit laboured, e.g., ‘Although I feel I am a member of my own culture…’, they are otherwise unremarkable in English. To the Japanese, references to one’s own and other people’s cultures have an unnatural and jarring clang. As one translator put it, “I don’t think Japanese people distinguish people by culture, but by country or nationality” (personal correspondence). In the Japanese worldview there are two types of people in the world, nihonjin (Japanese) and gaijin or gaikokujin (literally ‘outside’ people or ‘foreign-country’ people), i.e., non-Japanese. Finer cultural distinctions than that are generally not germane. It is difficult to say what effect the specific insistence on the word ‘culture’ (to which the Japanese version of the IDI remained faithful) might have had on respondents. One translator had to consciously make an effort not to substitute gaikokujin for the phrase ‘people from other cultures’ and was at times overcome by the unnaturalness of the construction to her Japanese sensibilities. It certainly made the items stand out, compelling the Japanese respondents to either reflect on the relationship in their minds between nationality and culture or to automatically substitute the more natural gaikokujin in their minds for ‘people from other cultures’. Another unintended consequence might have been to merely annoy respondents and weaken the ‘face validity’ of the instrument.

Conclusion

From the perspective of translation, the modified translation protocol field-tested in this study, which exceeded the existing ‘Gold Standard’, greatly enhanced the interpretation of the data yielded by the statistical and other components of the validity analysis of the IDI. The modifications consisted of the addition of two steps.

The most thoughtful approaches to translation (and adaptation) involve intensive consideration of the accuracy, naturalness, and acceptability of the translations. This, however, normally takes place within the translation context itself, circumscribing the focus and scope of the deliberations. One step that was added enhanced the evaluation of the translation because it occurred outside of the translation cocoon; using the draft translation as a rating tool for ‘amateur’ attempts to translate the items (step three of Table 1, above). The distinction might be subtle, but the deliberations that comprise the translator’s craft, options such as word choice, syntax, etc., are part of a self-referential exercise taking place within the context of the translation. Taking a step outside of the translation process to see how the items worked in the context of interpretations by the instrument’s intended final users led to different insights in terms of how non-translators might interpret wording or, more importantly, concepts.

The second contribution this study made, specifically to the translation of research instruments, was to introduce the step (step five in Table 1) of having a bilingual complete both versions of the instrument consecutively. Ideally, if the translation is ‘perfect’ and the instrument performs equally and equally well in both languages, the score for each item would be identical. Any discrepant scores would point to a problem either with the instrument or the translation. The score discrepancies uncovered by this addition to the protocol uncovered the existence of conceptual issues, as noted above.

The process of translation permitted an empirical investigation of claims that the IDI was suitable for use across cultures and languages and brought to light shortcomings that would otherwise have remained obscure.

References:

Bennett, M. (1986). Towards ethnorelativism: a developmental model of intercultural sensitivity. In Michael Paige (Ed.), Education for the Intercultural Experience (2nd ed.) Yarmouth, ME: Intercultural Press.

Bennett, M., & Hammer, M. (1998). The Intercultural Development Inventory (IDI) Manual. Portland, Oregon: Intercultural Communication Institute.

Behling, O., & Law, K. S. (2000). Translating questionnaires and other research instruments: Problems and solutions. Thousand Oaks, CA: Sage.

Edwards, J., & Dawson, D. (1983). My friend David: a sourcebook about Down’s Syndrome and a personal story about friendship. Portland, OR: Ednick Communications.

Greenholtz, J., & Morita, Y. (1988). Mai Frendo Deibido. Kyoto, Japan: Dohosha.

Hambleton, R. K., & Patsula, L. (1998). Adapting tests for use in multiple languages and cultures. Social Indicators Research, 45, 153-171.

Kelly, G. (1963). A theory of personality. New York: Norton.

Kristjansson, E. E., Desrochers, A., & Zumbo, B. D. (2003). Translating and adapting measurement instruments for cross-linguistic and cross-cultural research: A guide for practitioners. Canadian Journal of Nursing Research, 35 (2), 127-142.

Matsumoto, D., LeRoux, J., Iwamoto, M., Choi, J. W., Rogers, D., Tatani, H., & Uchida, H. (2003). The robustness of the Intercultural Adjustment Potential Scale (ICAPS): The search for a universal psychological engine of adjustment. International Journal of Intercultural Relations, 27, 543-562.

Sapir, E. (1929). The status of linguistics as a science. Reprinted in David Mandelbaum (Ed.), Selected writings of Edward Sapir. Berkeley: University of California Press, 1949.

Yamamoto, S., & Tanno, D. (2002). Ibunkakanjuseihattatsushakudo (The Intercultural Development Inventory) no nihonjin ni taisuru tekiyohsei no kentoh: Nihongo bansakusei wo shiya ni irete (Assessing the applicability of the Intercultural Development Inventory to Japanese: bringing in the viewpoint from the creation of a Japanese version). Journal of the Aomori National University, 7 (2), 24-42.

Zumbo, B. D., & Taylor, S. V. (1993). The construct validity of the extroversion subscales of the Myers-Briggs Type Indicator. Canadian Journal of Behavioral Science, 25, 590-604.

ACKNOWLEDGEMENTS

I would like to thank Drs Milton Bennett and Mitchell Hammer for granting me permission to translate the Intercultural Development Inventory into Japanese and to use it in my research.
I would also like to thank my research assistants, Vickie Yau, who kept all of my data in good order and was always there when I needed to be reminded how to find them, and Michi Ohashi, whose insights into language and cultural concepts helped shape my thinking about the issues involved in porting an instrument from one culture into another.
I am extremely grateful to my translation team, Makoto Morise, Yumiko Morise, Naoko Robb, and Shizu Yamamoto who contributed their expertise and their time, and to the raters, Masaki Kobayashi and Mitsunori Takakuwa whose insights helped me to refine and shape the final version.

——–

This report is based on Dr. Joe Greenholtz’ book Validity Issues In Translating Instruments Across Languages and Cultures: Exploring the Paradox of Differing Cultural Perspectives on Intercultural Sensitivity available on Amazon.

Dr. Joe Greenholtz
Phone/fax: (604) 241-2432
E-mail: principal@hjgconsulting.com
Website: www.hjgconsulting.com