Which of the following statements best describes what ethnographers refer to as an etic explanation?

Emic–Etic/Universal–Culture-Specific

The emic–etic distinction refers to the two traditional research strategies used to study phenomena in different cultures (Berry, 1989; Van de Vijver, 2010). The terms ‘emic’ and ‘etic’ were borrowed from the study of linguistics. Specifically, ‘etic’ refers to research that studies cross-cultural differences, whereas ‘emic’ refers to research that fully studies one culture with no (or only a secondary) cross-cultural focus. Etic research seeks to define common phenomena across cultures which can be used to define a set of universal phenomena among all cultures, whereas emic research lacks this aspect.

Proponents of the emic viewpoint posit that phenomena should be studied from within their own cultural context. For example, studies of Chinese students’ learning strategies should provide a detailed overview of these strategies and presumably relate them to relevant features of Chinese culture, such as Confucian philosophy (Watkins & Biggs, 1996). In contrast, etic researchers might use a questionnaire to compare the learning strategies of Chinese vs. American students.

Many researchers now recognize that the two strategies are better viewed as lying on a continuum rather than as mutually exclusive and incompatible (Cheung, Van de Vijver, & Leong, 2011; Helfrich, 1999). Thus, the emic–etic distinction has been redefined as referring to endpoints of a continuum that ranges from cultural specificity (emic) to universality (etic).

In this new approach, emic and etic aspects can coexist and work together because many phenomena studied in cross-cultural research have both universal and culture-specific aspects (Cheung et al., 2011). For example, despite its limitations (Boyle, 2008), there is evidence that the Five-Factor Model (FFM, comprising constructs labeled Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness) can be discerned universally. On the other hand, although the Western FFM may be found in all cultures, the model may not adequately capture culture-specific aspects of personality which may exist beyond the universal aspects of personality. Studies in China, for example, provide evidence that the FFM gives an incomplete rendering of Chinese personality; the model does not adequately measure the relational aspects of personality (e.g., the extent to which a person fits into his or her group, as a salient aspect of Chinese personality). Similarly, studies in various cultural groups in South Africa suggest that Agreeableness is too narrow a trait to cover all major social aspects of personality, such as softheartedness (involving aspects such as amiability, gratefulness, and empathy) and relationship harmony (involving aspects such as accommodating, flexibility, and humbleness; Nel et al., 2012). These concepts were frequently mentioned in semi-structured interviews when persons from all major ethnic groups in South Africa were asked to describe persons they know well. These interviews also revealed clusters of personality descriptions that could be captured in terms of the FFM; Extraversion, Neuroticism, and Conscientiousness could be easily discerned in the data, while Openness was found both in terms of intellect and broadmindedness.

Results and Comments

An emic/etic approach was used to develop the SAS. An inductive approach was used to identify the SAS factors and a deductive approach was used to revise the scale. Extensive research has been carried out to establish the psychometric properties of the scale. The SAS advances theory development in cross-cultural psychology by extending the domain of assessment from values to beliefs and from personal preferences to the social perception of the cultural reality.

Social Axioms Survey II

Basic debates never die: Emic vs. etic approaches

A key notion in crosscultural personality research is the distinction between imposed-etic (imported) and emic (indigenous) approaches to data collection (Berry, 1980). The imposed-etic approach, which is the most commonly used one, typically involves the use of translated instruments that were originally developed in another language and cultural context (e.g., McCrae & Costa Jr, 1997). Emic approaches, on the other hand, explore a particular psychological construct from within the culture. With the emic approach, instruments and theories indigenous to the target culture are developed by relying on a systematic process (through psycholexical studies, focus groups, interviews, content analyses of popular media, or culturally informed traditional scale development methods) that generates a set of local or “indigenous” attributes and stimuli (Saucier & Goldberg, 2001).

The imposed-etic approach, which is obviously much more economical, is appropriate when the researcher's main goal is to examine how the psychometric properties of a particular measure or culturally delimited construct generalize to other cultures or languages (i.e., measurement equivalence). For instance, a researcher may wonder whether, or to what extent, markers of the six NEO-PI-R facets of extraversion (gregariousness, assertiveness, energy, excitement-seeking, positive emotion, and warmth) and six facets of neuroticism (anxiety, hostility, depression, self-consciousness, impulsivity, and vulnerability) covary similarly and thus yield identical dimensions when administered in Spanish (the answer is yes; Benet-Martínez & John, 1998). However, the imposed-etic approach has a key limitation: its top-down nature assumes that the constructs under study (e.g., extraversion and neuroticism, or positive and negative emotionality) are defined in the culture of interest in exactly the same way as it is in the culture where the constructs and measure were developed (i.e., in terms of the six aforementioned facets for each dimension, no more nor fewer). In reality, local folk conceptualizations of the constructs might have some culture-specific (i.e., emic) elements that are likely to be lost when using translated instruments. For instance, emic lexical personality studies in Spain (e.g., Benet-Martínez & Waller, 1997) have shown that dimensions capturing pleasant vs. unpleasant affect (i.e., valence) and high vs. low arousal, rather than extraversion and neuroticism, best capture the affect components of personality. Most crosscultural comparisons that rely on translated instruments overemphasize similarity, precisely because culture-specific indicators of common constructs may have been missed (e.g., McCrae & Costa, 1997).

One obvious advantage of the imposed-etic strategy (besides its lower cost) is that it makes crosscultural comparisons across studies statistically feasible, given that quantitative judgments of similarity require stimuli that are equivalent. The emic approach, on the other hand, although more appropriate to identify culture-specific aspects of a construct (i.e., it has higher ecologically validity), renders empirical comparisons across cultures very difficult. What to do then?

One solution to the old emic-etic debate has been to pool both approaches into what is known as a combined emic-etic design (for a discussion and demonstration of this underused approach, see Benet-Martínez, 1999; Cheung, van de Vijver, & Leong, 2011). The application of a combined emic-etic approach involves the following steps: (a) identifying the possible emic (indigenous) elements of the construct (again, through focus groups, interviews, etc.) in the target culture and developing questions and items that adequately tap these constructs; (b) administering translated short measures of the construct (i.e., imposed-etic tests) in addition to the emic measures; and (c) statistically assessing the specificity and overlap between the imported and indigenous measures. This last step is key in that it allows researchers to objectively quantify how well imported and indigenous constructs overlap/differ and to clarify the meaning of the nonoverlapping indigenous elements (e.g., see Table 2 and Fig. 1 in Benet-Martínez, 1999). Another advantage of the combined emic-etic approach is that it allows for tests to examine whether the integration of imported and local elements into a construct has incremental predictive validity above and beyond that provided using either the emic or imposed-etic definitions alone (Church, 2001).

The combined emic-etic approach might seem very time consuming, and indeed it can be, particularly when the main goal is to systematically develop and validate comprehensive models and instruments for multidimensional constructs (e.g., the South African Personality Inventory, SAPI; Fetvadjiev, Meiring, Van de Vijver, Nel, & Hill, 2015).g But in other instances, this approach might mainly involve adding a few additional items and questions tapping into local definitions of a construct. Ultimately, flexible (top-down and bottom-up) combined emic-etic approaches that incorporate additional cultural elements into their instruments, when needed, might prove indispensable in an increasingly global and diverse world. Think, for example, of international corporations with a very international work force (e.g., Exxon, Google). In their personnel selection and evaluation efforts, these organizations might benefit from using instruments and conceptualizations (to tap into important work outcomes such as effort, loyalty, and resilience) that are sensitive to the meanings employees outside the US might ascribe to these personality traits (e.g., Schmit, Kihm, & Robie, 2000). With time, as these workers become fully integrated into the company's global workforce, these more culturally specific meanings might in fact become part of Western mainstream notions also.

IMPLICATIONS FOR FUTURE RESEARCH AND PRACTICE

The aim of this chapter was to provide a theoretical framework and specific propositions linking cultural context to the WFC phenomenon. The framework suggests that culture should be treated in two ways: first, as the main effect directly influencing WFC (its prevalence, type, direction) as well as the factors leading to WFC (i.e., the magnitude and type of demands and supports in work and family domains); second, as the moderator influencing the relationship of WFC with its antecedents and consequences.

The present review alludes to a number of tentative conclusions. First, cross-cultural differences in the appraisal and coping processes (Lazarus & Folkman, 1980) seem to account for the differences in the prevalence of WFC as well as its impact on life outcomes (e.g., psychological well-being, job and marital satisfaction, turnover intention). For instance, when employees appraise WFC as a threat rather than an opportunity for development, they tend to experience greater WFC. Similarly, when WFC is appraised as a life event harming the family, which is perceived to be the most important element in life, it is associated with greater health-related problems than when it is appraised as a sacrifice for family. Therefore, one important area for future research is cross-cultural differences in appraisal and coping processes.

Second, the review suggests that there are emic or indigenous manifestations of some of the key constructs used in WFC research. For example, in collectivistic and high power distant cultures, managerial support may include a paternalistic approach, which can enhance work-family balance. As another example, work demands in collectivistic cultures include the pressure to maintain harmonious relationships and avoidance of interpersonal conflicts at work. The stress created by this tension at work can lead to strain-based WFC. Even the simple term “family” may mean different things in different cultural contexts. Future research should examine the culture-specific or emic construals of demands and supports (i.e., what constitutes demand and support) in work and family domains.

Indigenous or emic perspectives allow us to unfold variform universals (i.e., general principles hold across cultures but the form or enactment of this principle varies; Bass, 1997) and variform functional universals (i.e., the relationship between variables is always found but the magnitude or direction may change depending on the cultural context). As discussed by Gelfand, Erez and Aycan (2007), indigenous perspectives not only contribute to the development of more universal knowledge, but also help us to understand our own culture and behavior. Pruitt states that:

… most cultural differences are relative, rather than absolute. In other words, people across the world are capable of behaving in almost any fashion, but their preferences for one kind of behavior over another differ from culture to culture. Characteristics that are dominant in one culture tend to be recessive in another, and vice-versa … By studying other societies where these features are dominant, they can develop concepts and theories that will eventually be useful for understanding their own (2004, p. xii).

Future cross-cultural research on WFC should invest more in emic or indigenous perspectives in order to provide a more universal science and to unearth recessive characteristics in other cultures.

Last but not least, this review lends itself to a number of tentative implications for policy and practice. As discussed at the beginning of the chapter, understanding cross-cultural differences in WFC has important implications for diversity management. First, organizations with a culturally diverse workforce should pay close attention to the implications of their career planning for work-family balance. Employees who value the balance between work and family more than the advancement in their careers may view the promotion to a managerial position not necessarily in a positive light. For them, the promotion may even be demotivating as it is deemed to be an unnecessary compromise from the family. Second, the nature and extent of organizational and managerial support to work-life balance may be evaluated differently by employees, depending on their cultural values. Those from collectivistic and high power distant cultures are more likely to expect organizational and managerial support and guidance in their private lives, whereas those from individualistic and egalitarian cultures would consider this as an invasion of their privacy. Managers should be trained to tailor-make their approaches in handling work-family problems of their diverse workforce. Organizational support systems or benefits and allowances can also be tailor-made to fit to different needs and expectations of the diverse workforce. A cafeteria approach could be adopted in benefits and allowances to support work-family balance to include various options to choose from, such as providing support to elderly family members, helping employees to find private tutors for their children, arranging for homemade meals to take home in the evening, and organizing visits or social events for the family members.

3.06.3.2 Phenomenological Approach

Consistent with an emic perspective, phenomenology is a method that aims to describe human experience as it is lived, felt, and understood by the participant (Wertz, 2011). Similar to grounded theory, phenomenology does not attempt to apply preexisting theories or frameworks to a phenomenon, but rather lets the phenomenon speak for itself, in all its ambiguity and as free as possible from any preconceptions or assumptions. Unique to phenomenology is its focus on conscious and direct lived experience, without attempting to derive meaning or causation beyond the subjective point of view of the participant. The phenomenologist investigates what is experienced by the participant, and how it is experienced (Wertz, 2011, p. 125). Whereas other qualitative methods often provide theoretical explanations for their findings, phenomenologists describe phenomena only from the point of view of the individual participant or group of participants (Wertz, 2011). Accordingly, phenomenology is particularly well suited for the in-depth exploration of Psychology concepts and constructs among individuals from various backgrounds.

Data analysis using a phenomenological approach involves an attempt to grasp what is essential to participants' psychological processes and experiences, and meaning-makings of these (Wertz, 2011; Husserl, 1962). Phenomenological data analysis usually starts with a holistic reading of an interview or focus group transcript, this in order to gain a sense of the narrative as a coherent whole. Second, the transcript is segmented into meaning units, each revolving around a particular theme or topic that is to be analyzed and reflected upon separately. Third, the phenomenologist proceeds to an in-depth analysis and reflection about the particular psychological significance of each meaning unit. Last, the phenomenological narrative (i.e., study report) consists of a comprehensive descriptive summary of a participant or group of participants' psychological structure as it relates to the meaning-making of the phenomenon of interest (Wertz et al., 2011; Giorgi, 2009). If the study involves more than one participant, descriptive summaries of individual cases are compared to determine which characteristics are essential, or invariably present, in each participant's narrative (Wertz, 2011).

Phenomenological data analysis also involves the use of free imaginative variation, which consists in allowing the analysis to go beyond empirically collected data to be compared, which would be limited to inductive generalization (Wertz, 2011; Husserl, 1962). In free imaginative variation, the researcher uses her imagination to reflect on whether a phenomenon, or characteristic, is possible (imaginable) in all conceivable examples (i.e., all individuals with characteristics similar to the participant in question). The goal of free imaginative variation is to gain general knowledge about the essential characteristics of a phenomenon (e.g., essential features of anxiety disorder among US college women; Wertz et al., 2011; Wertz, 2010; Husserl, 1962; Giorgi, 2009).

A recent study on the meaning of therapists' hope for their clients provides a compelling example of how phenomenology can be used to inform and improve clinical practice. Using data collected through semi-structured individual interviews of 8 psychotherapists, Bartholomew et al. (2019) used a phenomenological approach to understand therapists' meaning-making of hope for their clients, a phenomenon noted to be understudied despite having been identified as a central process in psychotherapy. Data analysis consisted in phenomenologically interpreting and coding meaning units that reflected participants' comments deemed relevant to the phenomenon being studied. For example, the statement “I'm noticing that I have that vision of positive improvements for them” was identified as a meaning unit and given the code “Ability to envision change as a recognition of having hope” (Bartholomew et al., 2019, p. 500). Once the researchers agreed upon a final list of meaning units, they began analyzing and grouping these into themes that captured critical aspects of the phenomenon of interest. Data analysis ended with a synthesis of themes that reflected the essence of therapists' lived experiences related to hope for clients. Results indicated that essential characteristics of therapists' hope for their client consisted of (1) a sense of holding and possibility; (2) fundamental, dynamic, and reflective practice; (3) client influence (positive and negative) on hope; and (4) connection through hope.

Cultural Factors in the Assessment of Personality and Intelligence

There are various examples of emic models of personality. These do not provide a pancultural view of personality, but try to describe personality in a specific cultural context as accurately as possible (Kim et al., 2006). For example, Doi (1973; see also Berry et al., 2011; Yamaguchi, 2001) has described a Japanese model of personality in which “Amae” plays an important role. Amae refers to “indulgent dependency” and the passive love and dependence that find their origin in the love of mother and child and form the template of relationships later in life. These studies are not comparative; as a consequence, bias and equivalence may not seem to be an issue. However, from a slightly more distant perspective, these studies address the bias in Western models, because they indirectly show where Western models of personality fall short. It is suggestive that in both models (as well as other non-Western models of personality) social and relational aspects of personality play a more central role than is common in Western models of personality.

A few Western-based personality inventories have been administered in many countries. A notable one is the Eysenck Personality Inventory. According to Eysenck, personality consists of three aspects: neuroticism, psychoticism, and extraversion. The inventory, used to assess these factors, shows these three factors with remarkable consistency. The factors underlying the instrument tend to be stable, indicating that the structural equivalence of the instrument, as tested by Exploratory Structural Equation Modeling, is good (Bowden et al., 2016). Another currently popular personality instrument is based on the Five-Factor Model of Personality. Extensive research has shown that the five factors that in this model are assumed to underlie personality (extraversion, neuroticism, agreeableness, conscientiousness, and openness), are remarkably consistent across cultures (Schmitt et al., 2007). However, studies of the measurement unit equivalence of the scales tend to produce more ambiguous results. Statistical tests (e.g., confirmatory factor analysis) do not provide clear support for measurement unit equivalence, but it is unclear to what extent this problem is due to the high sensitivity of the statistical tests employed to test the equivalence or to real differences in the measurement unit across cultures.

It can be concluded that cross-cultural research on personality has shown that the Western models can claim a broader generality than their cultural context of origin. Structural equivalence is strongly supported (it may be noted that studies have been restricted to schooled populations; it is still unclear to what extent these results also apply to illiterates). It is still unclear to what extent non-Western theories and models could add other universal factors. Recent studies suggest that the five-factor model of personality may be incomplete; researchers from Hong Kong found a sixth factor, interpersonal relatedness, in China; a similar factor was reported in the Philippines and South Africa (Cheung and Fetvadjiev, 2016). If new factors would emerge and it turns out that Western models are incomplete, it is likely that the new factors will be located in the social and relational domains.

The domain with probably the most widely documented cross-cultural research is intelligence. Standard (Western) intelligence tests have been administered in a wide variety of cultures; in addition, in many multicultural societies these tests have been administered to majority group members and usually at least one of the larger ethnic minority groups. Instruments such as Raven's Progressive Matrices and the Wechsler intelligence scales for adults and children have been found to reveal a remarkably stable factor structure across cultures, which confirms the structural equivalence of intelligence, as measured by these tests.

Statistical tests of measurement unit (metric) and full score equivalence of intelligence test scores tend to produce more ambiguous results. It is often difficult to decide whether equivalence in a given case is scalar equivalence or measurement equivalence. For example, the debate about ethnic differences in cognitive test performance can be largely seen as focusing on the level of equivalence of cross-cultural score comparisons and about the presence of biasing factors (notably method bias). For example, Rushton and Templer (2009) argue that when appropriate intelligence tests are used, cross-cultural differences in test performance reflect valid intergroup differences and show full scalar equivalence. The implication is that group differences in the intelligence test scores reflect differences in intellectual abilities while group differences mainly or exclusively reflect method bias in the reasoning of their critics. Many of their critics state that common intelligence tests show problems such as differential familiarity and that this method bias will only allow measurement unit equivalence.

Despite the disagreement about the higher levels of equivalence of cognitive test scores, cross-cultural studies show that it is very unlikely that intellectual processes are fundamentally dissimilar in different cultures. The findings are not supportive of theories that hold that different cultures produce their own kind of intelligence, but more supportive of theories according to which cultures can shape the same set of human cognitive processes in a differential way. However, recent cross-cultural work points to another possible one-sidedness of Western tests. These tests tend to focus on two aspects of intelligence: reasoning and memory. Implicit theories of intelligence have been investigated in a number of countries by asking participants which characteristics they associate with intelligence; social aspects are almost always found to be important; yet, social intelligence is not part of all intelligence tests. It is interesting to note that the social domain has also been mentioned in relation to personality research as an area in which expansion of Western models may be needed to accommodate non-Western models.

Thus, cultural factors impact on mental measurement in various ways. Some of these factors may be artifactual, such as a poor item translation. Other factors, however, are more basic and may refer to problems with the whole instrument or even the concept measured by the instrument. It is clear that methodological and substantive considerations cannot be separated when dealing with cultural factors in mental measurement. Knowledge of both mental measurement and of the culture to deal with is essential in establishing quality in cross-cultural mental measurement.

Emics and Etics

Participation can lead to a greater emic (insider's, subjective, viewpoint) understanding; observation tends toward an etic (outsider's, objective, viewpoint) understanding. The emic point of view is concerned with the meaning the people ascribe to events, relationships, behaviors, and experiences. The etic point of view attempts to map patterns of behavior using categories identified not by insiders but through the perceptions and interpretations of the “outsider” often based on preexisting theory. Theoretical insights are thus brought to the field rather than being grounded in or “discovered” through the field experience. A completely emic view is often only an ideal because the researcher brings his or her own political and moral values into the field, which at a minimum will result in (unconscious) selective perception and recording of phenomena. Furthermore, hypotheses and categories for observation are often developed prior to entering the field, although they are subject to revision. On the other hand, a reliable and valid etic understanding, to not be ethnocentric, must take into account the emic viewpoint(s), which may be heterogeneous. (Viewpoints differ, for example, by gender, class position, or any number of achieved or ascribed statuses.) As Peter Berger pointed out in 1976, people can be assumed to be experts in their own lives and experts in defining the immediate problems (or joys and satisfactions) that confront them. Sometimes they even describe the larger context of their day-to-day problems: Etic approaches are useful in analyzing this context but may have already been appropriated by at least some members of the community; for example, the knowledge that the profit-making propensities of the factories they work for are responsible for their low wages or that Green Revolution packages are not adapted to the local agricultural regime.

In combating ethnocentrism, the belief in and ethical principle of “cultural relativism” is endorsed by most participant observers. There have been modifications in the theme of cultural relativism—which tends to reify “culture”—with the growing recognition (i) that behaviors, practices, norms, and values within any one culture may be heterogeneous due to differential access to the means of production, power differences based on gender, ethnicity, and/or age and also the existence of subcultural groups; and (ii) that all cultures are and/or have been affected by world systemic processes (or processes of globalization), including colonialism, imperialism, capitalism, and consumerism.

5 Decentering

When cross-cultural researchers are sensitive to the emic/etic distinction, they are careful not to impose their own emics on other cultures. Many concepts in psychology (e.g., personality, intelligence, and worker motivation) reflect the culture in which they were developed and in which research on them took place. When one of the authors gave a lecture on personality to a group of Chinese scholars, they later commented that they found many of the specific terms (achievement orientation, need for power, openness to experience, etc.) difficult to understand since there were few readily available terms in Chinese. They mentioned that they have a much richer language to describe people’s relations with others, which would follow from China’s status as a collectivist culture. The key point is that personality terms in English and collectivist terms in Chinese may reflect the two culture’s emics. To expect an easy translation of them could be a mistake.

When researchers use decentering, they make attempts to find concepts that are readily expressible in the languages under study. No original language version becomes the “center” that has to be translated into less familiar and sometimes clumsy target language terms. Rather, the original language version is open to change as each translated/back-translated version is reviewed. With each review, researchers reword the original until a smooth target language version (as reported by translators or independent bilinguals) and back-translation version are developed. In the case of published standardized tests, this procedure demands that researchers be willing to revise the original wording just as if they were formulating an instrument for the first time. The assumption is that more meaningful information will be obtained from a decentered instrument than from a forced translation of English that has problematic equivalents in the target language.

Emic: ethnography and insider researcher

An ethnographic study suggests itself as a foremost example of emic considerations. Perhaps not as visible, but needing consideration nonetheless, is the insider researcher. A researcher may have phenomenology as the methodology, case study as the method, and participants as fellow practitioners in a given organization, but there is more to it in relation to insider. The insider researcher is actually involved with the participants as one of them, working alongside them, taking up the same goals and practices geared toward achieving them while researching these very things in which all are engaged…research. The insider researcher has a dual role, in effect, as both a participant and a researcher. This is not just a gulf between the profession and the laity, important as this is when discussing language, protocols, procedures, and conventions of a given profession. Rather than seeing this as a disadvantage, in qualitative research we take up its suggestive possibilities for our research.

10.01.4.4 Emics and Etics in Cross-cultural Research

I introduced above the idea of etic (universal) and emic (culture specific) cultural elements. Here I will elaborate on the meaning of these constructs and explain how they might be used in the measurement of constructs.

Emics, roughly speaking, are ideas, behaviors, items, and concepts, which are culture-specific. Etics, roughly speaking, are ideas, behaviors, items, and concepts which are culture-general—that is, universal.

Emic concepts are especially useful in communicating within a culture, where one word can sometimes be used to convey a very complex idea. For example, a geographer who has studied the inhabitants of the Tierra del Fuego in the Southern tip of South America, told me that they have a word, mamihlapinatapei, which means “looking at each other hoping that either one will offer to do something that both desire but are unwilling to be the first to do.”

One can almost see the scenario when boy meets girl and they mamihlapinatapei! One learns quite a bit about that culture by knowing the definition of this word. Similarly, by learning particular words, we get to know more about a culture.

Emic concepts are essential for understanding a culture. However, since they are unique to the particular culture, they are not useful for cross-cultural comparisons.

Let us consider an analogy, the comparison of apples and oranges. Apples and oranges have some common attributes, such as size, weight, price, and availability. They also have some unique attributes, such as unique flavor and aroma. Clearly, if we are going to understand what an orange is, we need to know about orange flavor. But we can not compare apples and oranges on orange flavor, except to say that apples do not have it. On the other hand, if we want to talk about price, we can compare apples and oranges. If we want to compare price to size or weight we can certainly do that, too. So, we can have a “theory” of size-weight-price relationships and see if it holds as well for apples as it does for oranges, and we can even extend it to other fruits. Emics are like apple or orangeflavor; etics are like size, weight, and price. So, for certain purposes, such as comparisons, we must use etic concepts; for other purposes, such as getting a real “taste” of the culture, we must use emic concepts.

More formally, emics are studied within the system in one culture, and their structure is discovered within the system. Etics are studied outside the system, in more than one culture, and their structure is theoretical. To develop “scientific” generalizations about relationships among variables, we must use etics. However, if we are going to “understand” a culture we must use emics.

Many anthropologists work with emics and think that etics are silly. They would say: “you do not know about apples by just knowing about their price, weight, and size.” Psychologists want to make generalizations about people, so they do not want to get into the details of a single culture. Cross-cultural psychologists try to both understand and compare cultures. They work with both emics and etics.

The important point is to find convergence between different methods of understanding reality. I do not agree with the view expounded by some “deconstruction” humanists, or even some who pretend to be scientists, who argue that nature is constructed, not discovered—that truth is made, not found (e.g., Haraway, 1991). I think that this is an extreme position which is not science, since it cannot be discomfirmed. On the other hand, I do grant that our subjective responses to reality are often constructed. I am a realist, so I do not believe that women are less good as Chief Executive Officers (CEOs), just because that is the male-dominated “construction” and conventional wisdom of current CEOs. But I would agree with this view if someone showed me a convincing study, with hard criteria, which indicated that this view is correct.

“Constructionism” can result too easily in an image of reality which is found in only one person's mind. It might be good literature, but it is terrible science. The essence of science is a conversation between scientists and nature. It requires probing with multiple methods, to replicate important findings, and establish convergence between observations and measurements. If such convergence is broad, that is, includes humanistic “findings” as well as findings obtained through scientific methods, we can be much more certain that we have identified an important phenomenon than if we have only one person's argument or a single set of findings obtained in one place, with only one method. An important distinction between the humanistic and the scientific method is the contrast between the subjective and the objective. If the humanistic insight is to be taken seriously it has to converge at some point with other evidence; it cannot remain entirely subjective.

If we take a construct generated in our culture and use it in another culture, we may have a pseudoetic (false etic) construct. We must get empirical evidence that the construct operates the same way in the other culture, that is, it is a true etic, before we use it to compare the cultures. Remember the discussion about construct validity above? That is what needs to be done to establish a true etic. Then we have to make sure that this etic construct is measured in ways that are culturally sensitive. That often requires the use of local terms and ideas. In short, we must use both etic constructs and emic ways to measure them. That is why cross-cultural psychologists advocate the use of both emics and etics (Berry, Poortinga, Segall, & Dasen, 1992; Triandis, 1972, 1992).

One advantage of such a strategy is that one can obtain more sensitive information about the relevant cultures, which implies that one can obtain more cultural differences if one uses both emics and etics in the design of cross-cultural studies. In one study (Triandis & Marin, 1983), samples of Hispanics and non-Hispanics responded to either a questionnaire based on ideas generated in previous studies with American samples (pseudoetic) or a questionnaire which used ideas obtained from focus groups (informal discussion groups which generated items to be used in the questionnaire) of Hispanics and non-Hispanics. The latter questionnaire included both emic items which were spontaneously generated only by the Hispanic focus groups and etic items which were generated by both the Hispanic and non-Hispanic focus groups. Since the questionnaires in that study had 600 questions, one could get six cultural differences by chance (at the 0.01 level of significance). The number obtained by the pseudoetic questionnaire was 14 and the number obtained by the emic plus etic questionnaire was 50. In other words, one can miss a lot of subtle information about the way cultures differ if one does not use emics.

The importance of emics cannot be overestimated. There are emic concepts that are extremely difficult to understand by people who use an etic framework.

Consider this example: What is a geisha? Most Westerners are likely to freely associate “prostitute” with that concept, and as a result they will be quite wrong. A 500+ page ethnography about geishas (Dalby, 1983) shows that the closest association to “geisha” should be “jester.” Just like the jester in a king's court had the function of diverting the king and his guests, so a geisha has as her chief function entertaining the clients of an establishment. Most geishas spend their time reciting poetry, singing, dancing, and serving their clients while they eat and drink.

Furthermore, there are many kinds of geishas, and only a minority are actual prostitutes (yujo). Most do other things, including greeting people at the door or playing musical instruments. The most highly sought geishas are older, more experienced, and thus able to act more appropriately. President Ford, the first US President to visit post-war Japan, had dinner with several geishas as part of his touring experience. The most beautiful were placed next to him; but the ones who did most of the entertaining were the older, more experienced ones. Of course, without knowledge of Japanese, President Ford may not have been able to appreciate the entertainment.

The contrast between “geisha” and “wife,” in terms of the free associations of these words among Japanese males, is instructive. The former is seen as sexy, artistic, witty, and economically self-sufficient; the latter as sober, humdrum, serious, and economically dependent. The former is likely to be well read, know the latest poems and songs; the latter is likely to talk about the problems of the children.

Now that you have a glimpse of the meaning of geisha, do you see that it is a culturally specific role, that corresponds to an emic Japanese term, and the Western view that she is a “prostitute” is a pseudoetic, and becomes a true etic only when it uses the idea of “jester?” The general point is that using our own emic terms to understand other cultures will generally result in misunderstandings. We need to obtain etic terms which correspond to the emics.