Tuesday, August 13, 2024

Translations: Lessons learned for better outcomes

Translating documents, advertisements or other materials can be a real headache for many organizations, especially with periodic news stories reporting on the explosive growth of groups, like Asians and Latinos. For example, one recent report by the Census Bureau cited in Texas Tribune tells us that Asians now have a higher population growth rate in Texas (5.5%) – faster than the overall population growth rate for the state of 1.6 percent.[1]  Latinos, however, continue to fuel a majority of the population growth in Texas and many U.S. communities.  The projected population growth in Texas is vividly illustrated by Figure 1 below. [2]


 Figure 1 Projected Texas Population by Race/Ethnicity

  












According to these projections of the Texas population by the Texas Demographic Center (January 2019), the period between 2020 and 2050 will realize a dramatic multicultural population boom of 108.9 percent for Non-Hispanic Blacks, 509.7 percent for Non-Hispanic Asians, and 113.4 percent for Hispanics – indeed, a significant change that is likely to  significantly increase the demand for language translation services. The same report revealed that four of the top ten counties in Texas that are projected to absorb this growth are located in the Dallas/Ft. Worth metro area (see Table 1 below):  Collin County (1.7 million persons); Denton County (1.7 million persons); Dallas County (1.4 million persons); and Tarrant County (1.4 million persons).  If your organization serves or plans to serve Asian and Latino persons with your services or products, a sound translation strategy will be especially important. 

Indeed, one medical institution recently contacted me with an urgent question:  With so many Asian groups in Dallas County, which languages should be targeted for translation?  This sense of urgency is especially acute in industries related to healthcare, legal, and public safety where even minor errors in translation can have serious consequences. For example, communications and medical interventions in linguistically diverse U.S. communities during the Covid-19 pandemic were compromised by the absence of information regarding the racial-ethnic identity and language characteristics of highly impacted communities.    Indeed, about 22 Asian countries are represented in North Texas by the Census Bureau while Latinos are represented by 21 countries – numbers that clearly underscore the need for translation support.  Moreover, Asian countries include hundreds of languages with different dialects.[3]  And while Spanish is common to Spanish-speaking countries, their dialects are known to vary across these countries.[4]

 Table 1: Top 10 Texas Counties with Greatest Numeric Change between 2010 and 2050

 

 The translation strategy that an organization adopts, however, can produce negative outcomes and undermine community goodwill if not done carefully and professionally.  To clear up some of the confusion associated with translations, I would like to share some lessons learned from my 45 years of practical experience in conducting surveys and experiments with Latinos and Asians for a broad spectrum of private, public and academic clients.  While I am not a linguistics expert, I have studied the use of language in these studies, and this experience has been reinforced by teaching undergraduate and graduate courses in Hispanic marketing, multicultural research methods, survey research methods, statistics, and mass communications. Thus, the lessons learned are not just subjective impressions about language usage but grounded in academic and real-world experiences. Although I have addressed language translation issues in the past, recent events pointed to the need for an update on this topic.

 Lesson 1:  The language that Latinos and Asians use to communicate is not uniform and shaped by key characteristics such as age, nativity, income and education. 

For example, a non-English or foreign language is more likely to be used by persons who are foreign-born, older, less educated and lower income. Persons who are are more likely to communicate in English include persons who are native born, younger, higher income and more educated groups. Language usage is also likely to vary in terms of the context:  reading a book, listening to a medical provider, watching television, or talking with friends or family members.  Consequently, prior to initiating a translation process, it is very important to identify the specific target audience within the Latino or Asian household or communities that the translated document is designed to benefit. A good rule of thumb is to provide both English and native-language documents to ensure adequate coverage of all members of a household.

 Lesson 2:  A head nod is not sufficient to confirm understanding of a translated document.

 It is often the case that a Latino or Asian person is presented a form to complete a transaction or register for a program or service.  Attorneys, healthcare providers, and mortgage companies often present documents that need signatures on documents that involve important decisions. In such circumstances, the customer is typically asked if they understand what is being explained or what they are reading, and a head nod or “yes” response is usually accepted as confirmation that the information was “understood.”  But is this type of confirmation a valid one?   Perhaps not if there is no follow-up to determine that the information was really understood.  Latinos and Asians, especially immigrants, will often confirm that they understand an instruction or a document to save face and not admit that they lack the ability to read or understand that language. Consequently, it is important to ask these customers to explain or demonstrate their comprehension of an instruction by asking them to repeat in their own words what the instruction means or asking them to physically demonstrate their comprehension. In a healthcare setting, for example, it would be advisable to ask a Latino or Asian to repeat, in their own words, the instructions for taking their medication(s) – clearly a more valid measure of comprehension than a simple head nod.

 Lesson 3:  Translators are not the last word on language translations.

 Over the past years, I have used translators for a variety of tasks using different languages and value the function that they serve. In one national study of Latinos, African Americans and Asians, our firm conducted surveys in five languages (English, Spanish, Chinese, Korean and Vietnamese) using mail, online and telephone survey methods – quite a challenge indeed. Over the years, I have learned to appreciate the value of a certified translator with experience in the subject matter at hand, whether legal, healthcare, insurance, etc. – which helps to reassure clients about the accuracy of the translation. However, it is a mistake to think that your job is done when a translator submits their finished product.  In addition to the accuracy of the translated document, it is equally important to know who the intended audience will be and their ability to read and understand the document.  Translators do not always know who the intended audience is and sometimes produce documents whose reading difficulty level is too high or include words or phrases that are unfamiliar to the intended audience. Mandy Sha, a Chinese research colleague of mine, was part of a research team that conducted a usability test of a Census Bureau questionnaire with Chinese respondents who could read and write in Chinese but had limited English-language proficiency. The investigators discovered several problems with the translated questionnaire, such as sentence structure or verb tense, grammar, lack of context, name order, validation rules and navigation instructions. [5]  By conducting a copy test, pilot test or usability study of a translated document, one can have the added assurance that the appropriate communication has been established with the targeted consumer.    Hence, your target audience, not the translator, should have the last word on the acceptability of a translated document or other material.

 Lesson 4:  Speaking a language is not an automatic qualifier for reading or writing in that language.

 Naïve marketers are often surprised to observe that Latinos or Asians can be conversing in their native language quite comfortably but may have difficulty when asked to read or write a document in that language. What some marketers fail to understand is that a language usually includes four basic functions --  reading, speaking, writing, and listening --  and proficiency in one of these functions does not necessarily mean proficiency in the other functions. In addition, many immigrants coming from Asian or Spanish-speaking countries may lack formal education and cannot read or write in any language, while more educated immigrants may be more literate in their native language.  By better understanding the origins and educational background of Latinos and Asians, marketers can develop communications that will be understood by their target audiences. Thus, one should always consider the language function being utilized when evaluating a translated document. It is insufficient to just ask a Latino or Asian person in which language they prefer to communicate since it is important to know whether their language preferences are dependent on the language function (i.e., reading, speaking, writing, listening) and the specific task or activity that they are engaged in (i.e., watching television, reading a book, talking with family, listening to music, conversing with medical staff).  So, learn to be situation specific in asking Latinos or Asians which language they prefer for communications.

 Lesson 5:  While the U.S. Census Bureau collects language usage data about Asians and Latinos, it can be misleading.

 Organizations often quote language statistics collected by the Census Bureau as evidence about the number of native-language households that reside in the U.S. at any point in time.  The quality of this language data, however, is limited in several ways.  First, one question in the American Community Survey asks:  “Does this person speak a language other than English at home?” If the question is answered “yes,” then the next question that follows is: “What is this language?  Thus, we learn from these two questions the number of persons that speak Spanish or another language.  But it does not ask how well Spanish or other non-English languages are spoken – referred to as language proficiency.  Thus, if a person utters one word in Spanish, then they are classified as a Spanish speaker or Spanish-language household. How well they speak Spanish does not seem to matter here. 

English is the only language where proficiency is asked:  “How well does this person speak English?” – to which one is provided four options:  “Very well,” “Well,” “Not well,” and “Not at all.”  While this type of language proficiency question is useful in providing some guidance on how well a person speaks the English language, other research that I have summarized elsewhere shows that Latinos tend to over-estimate their language skills on self-reported measures like the one used by the Census Bureau – a consequence of social desirability.  That is, native-born Latinos who are more English-language proficient often want others to think that they speak Spanish better than they actually do.  Immigrants, who are more Spanish-language proficient, often want others to think that they speak English better than they actually do.  Even when they claim proficiency in both languages, most native-born Latinos will choose an English-language interview when given a choice, while most Latino immigrants will choose a Spanish-language interview.  Hence, the language that a Latino chooses when provided a choice is a more valid indicator of their language dominance than their self-reported language proficiency.  Our experience suggests that Latinos should always be provided the choice of English or Spanish when asked to complete a task – such as an interview or a written document. This simple procedure will usually assure that you will get a more valid response.  Thus, the Census Bureau language questions, while useful, are crude measures of language behavior that should be used cautiously when evaluating the language behavior of U.S. Latinos or Asians.

 Lesson 6:  Using Latino or Asian employees to translate documents is risky.

 As a shortcut, some companies will utilize Latino or Asian employees to translate documents or interpret on the spot when the situation demands it.  Unless you know the training and education of that employee, you are taking unnecessary risks in assigning them this responsibility. Latinos or Asians who are born in the U.S. rarely study their native language formally in school and rely on the native language that they have heard or used growing up in their communities – often a mixture of English and a native language.  Important documents that relate to personnel procedures, healthcare, safety, insurance or legal matters should only be translated by a certified translator and evaluated by a usability test to ensure that persons in the target audience understand the translated documents or other visual aids.  Experience also suggests that graphic symbols, such as those used in hazard warning signs, also have cultural components that may not communicate the same message to culturally diverse groups. Copy or usability testing is especially important with exhibits that include hazard or warning signs that often rely on graphic symbols since they are used to warn or prevent injuries or accidents. Employees who live in the local area and have some familiarity with a native language can be helpful in identifying informal words or phrases spoken by local Latinos or Asians instead of more formal words or phrases.  For example, automóvil is the formal word uses for automobiles although the word “carro” is often used by some Mexican residents.  

 Lesson 7:  Knowing a language does not necessarily mean that you know the culture.

 In the employment world, many occupations require proficiency in one or several languages.  While proficiency in a language other than English is a definite asset in many jobs, it should not be confused with knowledge of a particular culture.  It is not uncommon, for example, for a foreign-born Latino or Asian with an excellent command of their native language to receive more consideration for a job than a similarly educated Asian or Latino whose native language proficiency is not as well polished --  the assumption being that a higher language proficiency also means more knowledge or experience with the culture.  This assumption may not necessarily be a valid one since a native-born Latino or Asian may indeed have more knowledge of the U.S. Latino or Asian culture than a foreign-born Latino or Asian who happens to communicate well in a native language. If the job involves responsibilities with U.S. Latino or Asian consumers, then knowledge of the U.S. Latino or Asian culture should be just as important in employment decisions as proficiency in a language.

 Lesson 8:  Language ability depends on sight and sound as well.

 My experience in conducting focus groups with Latinos and Asians has shown that their ability to understand a written document or verbal instruction is sometimes influenced by limitations related to visual acuity and hearing impairments.  Indeed, past research on vital statistics reveals that Latinos (11.2%), Blacks (10.0%), and Asians (9.7%) have trouble hearing, while vision limitations were reported for Blacks (9.3%), Latinos (9.4%), and Asians (5.7%). [6] Persons with vision or hearing limitations will not readily admit when they are unable to see very clearly, but an astute observer will notice non-verbal cues that suggest a vision problem.  Similarly, hearing impairments can be subtle and not usually something that will be readily apparent.  As a moderator, I have addressed such issues by reading a document out loud so that everyone can hear and understand the instructions and ensure that any written documents are provided in large fonts to enhance their readability. Rather than embarrass a person because they cannot see or hear very well, it makes more sense to offer options that will allow all persons to participate in the task or activity. 

 Lesson 9:  Is your target audience really diverse? How many languages need to be translated?

Occasionally, I have received calls asking how many languages need to be included in translating important documents, especially as it concerns Asian communities.  Because the Census Bureau data clearly shows “wide diversity” in some communities that represent multiple Asian countries, it becomes perplexing for some decision makers to determine which languages should be targeted for translation.  The solution, however, is rather simple.  The Census Bureau provides tables that show the country of origin represented in any community by geographic areas in the U.S. and its territories – currently there are 22 Asian countries and 22 Spanish-speaking countries counted by these Census tables. After consulting these tables, for example, one can just focus on translating a document for the countries that account for a majority of the population of Asian persons.  Although Spanish would be the only relevant language for the 22 Spanish-speaking countries, one can expect variations in dialects, idioms, and vocabulary depending on the concentration of these groups in any community. Knowing that a community is mostly Mexican, for example, might be a consideration for the translator to utilize idioms or vocabulary that are more commonly used by Mexican-origin persons.

 Lesson 10:  Using accents in Spanish translations or diacritical marks in Asian languages is essential.

Accents in Spanish-language text and diacritical marks in Asian languages are important linguistic features that help to guide the reader in the proper pronunciation and interpretation of a word or phrase.[7]  Without these linguistic features, foreign-language speakers are likely to mis-pronounce or misinterpret words or phrases.  Special keyboards are typically used to produce accents or diacritical marks. A translated document that lacks these linguistic features suggests that the translator was not adequately trained or lacked the proper tools to complete this task. You should avoid sharing translated documents with Latino or Asian persons without the required accents or diacritical marks. 

Lesson 11:   Language gets more complicated for messaging in advertisements.

If your translation efforts are focused on advertising messages, you will be faced with additional challenges that will likely exceed the skill set of a translator since this is not their area of expertise.  Indeed, you will need the help of a qualified advertising or public relations professional that has been trained in this area.  You will now have two areas of concern --  language translation and persuasive messaging – which is likely to vary considerably depending on the demographic subgroups that you are targeting.  For example, younger Latinos or Asians will likely be more responsive to an English-language advertisement although it has become popular for youths to mix English and a native-language in their communications. Older Latinos or Asians will likely be more responsive to advertisements that do not mix the languages which could be perceived as an indifference to proper grammar in that language.  To ensure the acceptance and approval of an advertisement, you should plan to conduct focus group copy testing of these ads before they are launched with the appropriate target audience to remove any uncertainties.

Lesson 12:  Providing language options also helps to build rapport.

 I have learned from past experiences that some organizations consider it burdensome to provide support in a language other than English to their customers – citing such issues as additional costs related to translators, bilingual staff, printing, advertisements and customer support.  However, one important benefit of providing support for non-English languages is the rapport that is established with Latino and Asian persons.  In our survey practice, for example, cooperation rates among Latino or Asian respondents are often improved when they are given the choice of a native language or English language interview – especially if the interview includes a foreign-born person.  Improved rapport not only leads to higher cooperation rates, but often results in the successful completion of the entire interview and better quality data that results from better comprehension of the survey questions. With higher survey response rates, the additional costs for support a language other than English are often justified.

 Note:  I welcome your feedback on the post.  Please send all comments or requests for a pdf version to edward@rinconassoc.com


Reference Notes


[1] Garcia, B.  (June 2024).  Fastest growing group in Texas:  Asian Americans.  Texas Tribune, Accessed at: https://www.texastribune.org/2024/06/26/texas-asian-american-growth-census/ 

[2] Potter, L., Valencia, L., Robinson, S. (January 2019). Texas Population Projections 2010 to 2050, Texas Demhttps://en.wikipedia.org/wiki/Languages_of_Asiaographic Center.  Accessed at:  https://www.demographics.texas.gov/

[3]  Wikipedia (2024). Languages of Asia,  Accessed at: https://en.wikipedia.org/wiki/Languages_of_Asia

[4] (August 2024).  Spanish dialects and varieties.  Wikipedia.org, Accessed at:  https://en.m.wikipedia.org/wiki/Spanish_dialects_and_varieties

[5] Wang, L., Sha, M. and Yuan, M. (2017).  Cultural fitness in the usability of U.S. Census Internet survey.  Survey Practice, Vol. 10, No. 3, 2017.  Accessed at www.surveypractice.org

[6] Rincon, E.T.  (2020).  The Culture of Research.  Publisher:  Writers Marq LLC.

[7] Thesaurus.com (April 2022) 12 Types Of Diacritical Marks And How To Type Them, Thesaurus.com, Accessed at: https://www.thesaurus.com/e/grammar/diacritical-marks/



No comments:

Post a Comment