英语西班牙语和葡萄牙语外文文献

An update on the roles of culture and language in designing emotion lists English Spanish and Portuguese Hannelize van Zyl a Herbert L Meiselmanb aHEINEKEN Burgemeester Smeetsweg 1 Zoeterwoude 2382PH The Netherlands bHerb Meiselman Training and Consulting Services P O Box 28 Rockport MA 01966 USA a r t i c l ei n f o Article history Received 12 November 2015 Received in revised form 24 February 2016 Accepted 25 February 2016 Available online 27 February 2016 Keywords Emotion Culture Language Beverage a b s t r a c t This paper is an update on a study investigating the effect of culture and language on reported emotions and confi rms that both culture and language are important in studying emotions evoked by beverages In addition sub categories of products such as beverage types have different emotion associations in differ ent cultures A list of emotions developed for beverages in the UK US Mexico and Spain was translated into Portuguese for Brazil and Portugal An on line questionnaire combining the emotions selected by the focus groups was then completed by 600 respondents each in Brazil and Portugal where people were asked which emotions applied to their favourite beverage beer and their least liked alcoholic beverage Data from the two studies were combined Respondents from English speaking countries showed very similar emotional reactions for individual beverage types Respondents from Mexico and Brazil were more similar to English speaking respondents than to those from Spain and with the exception of wine respondents from Portugal were similar to those of Brazil and Mexico in their emotional reactions Respondents from Spain and Portugal were sim ilar in their emotional reaction to wine We confi rm our conclusion that for products but certainly for beverages culture will affect emotion language usage but even within the same category the pattern of differences may be different for different products In designing emotion lists it is therefore important to take both culture and language into consideration and to realize that a list developed in one country for a specifi c product type is not necessarily suitable in another country or for a different product 2016 Elsevier Ltd All rights reserved 1 Introduction In a previous study van Zyl Portuguese is also spoken by more people in America than in Portugal Both Spanish and Por tuguese are among the 6 most commonly spoken languages in the world Wikipedia 2 Method In the fi rst study by van Zyl and Meiselman 2015 a list of 66 terms was compiled from a combination of previously published http dx doi org 10 1016 j foodqual 2016 02 019 0950 3293 2016 Elsevier Ltd All rights reserved Corresponding author Food Quality and Preference 51 2016 72 76 Contents lists available at ScienceDirect Food Quality and Preference journal homepage lists and focus groups in 4 countries and two languages In this follow up study the aim was to reach a balance between starting the entire process from scratch with focus groups in each country or simply translating an existing list It was decided to translate the long list compiled from focus group results in English and Spanish and then to use on line research in Brazil and Portugal to identify which of those words are most appropriate for each beverage type in these two Portuguese speaking countries The decision was based on the fact that the words from the English and Spanish focus groups were not vastly different for beverages in general and that the differences found between countries were related to their use for specifi c beverages in the on line study The list of terms was translated into Portuguese as used in Portugal and Brazil respectively resulting in 69 word pairs when combining the Eng lish Spanish and Portuguese lists The 69 word pairs are a result of words with similar meanings As one English word sometimes had two Spanish counterparts e g Friendly Simpatico and Amable and two Spanish words sometimes had one English counterpart e g Cari oso Loving and Affectionate there was a total of 69 word pairs Thus Friendly Amable is one word pair and Friendly Simpatico isanotherwordpair whileLoving Cari osoisapairand Affectionate Cari oso is a pair Each respondent only saw terms in their own language 66 and the word pairs were only used for the purposes of data analysis 2 1 Internet term selection The objective was to determine which of the 66 terms were most relevant to beverages for Portuguese speaking respondents with the main focus on which emotions one should study for bev erages that are liked rather than emotions related to a specifi c bev erage type On line research was conducted with 600 respondents each in Brazil and Portugal Respondents were recruited in three age groups legal drinking age 29 30 45 and 46 55 with 50 male and 50 female in each All respondents were specifi cally recruited to be native Portuguese speakers and drinkers of several types of beverages As in the previous study questions were posed to respondents sequentially and it was not possible to return to earlier questions to modify responses First respondents were asked to imagine hav ing their favourite beverage There were no instructions regarding the type of beverage or frequency of use While this was not stated in the questionnaire it should be kept in mind that a favourite bev erage is not necessarily the one consumed most often Respondents were asked to indicate for each of the terms whether their favour ite beverage made them feel more like that less like that or if the term was not applicable Respondents were not asked to scale the intensity of the emotions they experienced as the objective was not to compare beverages In the next question respondents were asked to indicate whether or not the beverage contained alcohol and then to choose from a list to identify the beverage they imag ined Subsequently respondents were asked to imagine that they were having a beer and to indicate for each of the 66 terms whether beer made them feel more like that less like that or if the term was not applicable There was no instruction regarding the type or brand of beer leaving respondents free to imagine their favourite style and brand of beer Respondents who had already identifi ed beer as their favourite beverage were not asked this question to avoid duplication In the fi nal part of the questionnaire respondents were asked to identify their least liked alcoholic bev erage from a list Respondents who had identifi ed beer as their least liked beverage were not asked the last question as they had already identifi ed their emotions related to beer All other respon dents were asked to imagine having their least liked alcoholic bev erage and to indicate for each of the 66 terms whether that beverage made them feel more like that less like that or if the term was not applicable Table 1 Distribution of favourite and least liked beverage as identifi ed by respondents Australia New Zealand UK USA Spain Mexico Portugal Brazil Favourite beverage Alcohol free beer0 20 80 50 54 21 00 81 5 Alcopops1 00 23 50 32 32 30 54 3 Beer25 925 327 229 231 434 235 648 3 Bourbon8 611 51 52 01 00 20 00 0 Cacha a 1 20 7 Cider4 02 010 00 70 70 51 80 3 Coffee9 610 87 35 47 08 17 05 8 Flavoured water0 20 50 20 00 02 70 50 2 Gin1 52 22 22 63 10 72 00 0 Juice2 73 82 82 34 76 56 37 8 Other6 68 54 77 65 93 28 51 7 Soft drink soda7 03 85 311 114 75 36 83 2 Tea5 16 28 24 54 22 86 01 0 Tequila1 21 20 54 11 37 30 00 8 Vodka6 66 87 210 14 96 04 74 0 Water2 00 51 23 12 61 54 32 0 Whisky6 34 73 55 16 010 83 24 0 Wine11 611 314 311 46 07 110 714 3 Least liked beverage Alcopops10 010 210 58 86 913 86 02 5 Beer11 510 49 011 44 612 17 56 3 Bourbon7 810 48 09 810 18 65 73 5 Cacha a 16 732 0 Cider10 38 35 78 17 213 85 07 0 Gin16 816 517 213 115 28 511 010 0 Other8 06 26 58 83 94 14 36 5 Tequila9 310 410 912 615 47 15 25 3 Vodka6 86 05 77 314 113 88 76 5 Whisky12 312 221 913 119 011 124 217 0 Wine7 39 54 77 33 87 15 73 3 H van Zyl H L Meiselman Food Quality and Preference 51 2016 72 7673 Data from Portugal and Brazil were combined with that previ ously obtained in English and Spanish speaking countries and results are reported for favourite and least liked beverages The dis tributions of favourite beverages and least liked alcoholic bever ages are presented in Table 1 Data was analysed for favourite beverages with and without alcohol As the percentage of respon dents who chose a beverage as their favourite differed among countries and as beverage type could be expected to have an effect on responses data was also analysed by beverage type In the case of favourite beverages as cell sizes for individual beverages were too small in some instances beverage types were grouped together as follows cold beverages without alcohol hot beverages without alcohol beer spirits and wine Other beverages such as alcopops and cider were excluded due to small cell sizes and not fi tting within the grouped categories The category other was excluded as it gave no information on beverage type This resulted in 4338 respondents from eight countries In order to show differences between countries and beverage types a covariance PCA was performed on the percentage of respondents who indicated that their favourite beverage made them feel more like this for all positive emotions and less like this for all negative emotions The analysis was performed both by grouped beverage types and also for beverages with and with out alcohol For the least liked alcoholic beverage a covariance PCA was performed on the percentage of respondents who indi cated that the beverage made them feel less like this for all pos itive emotions and more like this for all negative emotions 3 Results and discussion Fig 1 shows the observations plot for the PCA for grouped bev erage types Proximity of country beverage pairs in the PCA plot indicate that respondents from those countries reacted in the same way to their favourite beverage The further away country bever age pairs are from each other on the plot the more different respondents from those countries were in their reactions to a specifi c beverage There are two main results The fi rst is that respondents from Brazil reacted similarly to those from Mexico Respondents from Mexico and Brazil were similar to English speak ing respondents along factor 1 which explains most of the variation in the data but shows some separation from English speaking respondents along factor 2 The second main result is that respondents from Portugal reacted similarly to those from Mexico and Brazil for most bever age types with the exception of wine where their emotional response was similar to those of respondents from Spain on both factors 1 and 2 This is an interesting fi nding which should be explored further Other researchers have shown a differentiation Fig 1 Observations plot of covariance PCA on the percentage of respondents who indicated that their favourite beverage made them feel more like this for all positive emotions and less like this for all negative emotions analysed by beverage type 74H van Zyl H L Meiselman Food Quality and Preference 51 2016 72 76 between respondents from different countries with regards to their associations with wine Mouret Lo Monaco Urdapiletta and Parr 2013 found that New Zealanders considered wine a subject of relaxation and fun but also linked to food while the French asso ciated wine with friendship red wine and cheese Mora and Moscarola 2010 found that French consumers asked to write a narrative regarding wine consumption focused mainly on sensory aspects places and value judgments for example opinion and debate while Charters and Pettigrew 2008 found that unwinding and relaxation were important aspects of wine across a range of Australian respondents As before respondents from all countries found both cold and hot beverages without alcohol to be different from beverages with alcohol regarding the emotions they evoke Fig 2 shows the obser vation plot for beverages with and without alcohol In this PCA the differentiation found for wine is lost in beverages with alcohol between Portugal on the one hand and Mexico and Brazil on the other as wine was one of several alcoholic beverages regarded as a favourite beverage Regarding beverages in general respondents from Portugal Brazil and Mexico showed high agreement with each other as can be seen from their proximity on the PCA plot for most beverages with the exception of wine for Portugal As in the previous research there are fewer groupings of coun tries and beverages for least liked alcoholic beverages Fig 3 than for favourite beverages but there are some interesting trends For wine and alcopops respondents from Spain and Portugal reacted in a similar way which differentiated them from the other 6 countries along factor 1 of the PCA However in the case of cider Portugal and Brazil were grouped together and were separated from the other countries along factor 1 In terms of beer there was very little separation between countries along factor 1 but Mexico Spain and Brazil were grouped together and separated from Portugal and the English speaking countries along factor 2 The distribution for spir its was more homogenous than for other least liked beverages with only a small separation of Spain and Portugal from the other 6 countries This higher degree of similarity between countries for least liked beverages is to be expected as these beverages do not have an emotional benefi t but rather a common associa tion namely to be avoided resulting in a closer emotional association These results confi rm that culture plays at least as important if not more important a role as language in the use of emotion terms and that even for one beverage type an emotion list suitable for one country might not be suitable for another While these fi ndings are limited to beverages which were the subject of this research it is possible that the recommendation to not generalize emotion lists will also be relevant to other categories of products including both food and non food products Fig 2 Observations plot of covariance PCA on the percentage of respondents who indicated that their favourite beverage made them feel more like this for all positive emotions and less like this for all negative emotions analysed by beverages with and without alcohol H van Zyl H L Meiselman Food Quality and Preference 51 2016 72 7675 In our previous research respondents from English speaking countries showed very similar emotional reactions for individual beverage types The combined results show that respondents from Mexico and Brazil were more similar to English speaking respon dents than to those from Spain and with the exception of wine respondents from Portugal were similar to those of Brazil and Mex ico in their emotional reactions Respondents from Spain and Por tugal were similar in their emotional reaction to wine This further research confi rms our previous fi nding that for products and certainly for beverages culture will affect emotion language usage Even within the same product category the pat tern of differences in emotion language may be different for differ ent groups of products In designing emotion lists it is therefore important to take both culture and language into consideration and to realize that an emotion list developed in one country for a specifi c product type is not necessarily suitable in another country or for a different product even when the two countries share the same language However from a practical point of view the pro cess of generating emotion lists from scratch for each country is both time consuming and expensive Within a region it might be possible to translate a longer list from existing research before nar rowing it down for a specifi c languag