Colombians’ Sociocultural Representations of French and U.S. Citizens Delving into Intercultural Perceptions

This study analyzes sociocultural representations about French and U.S. citizens. Drawing on a word association methodology, a thematic and semantic analysis was conducted on a corpus of 4000 words, obtained from surveys completed by 200 Colombians. Findings indicate that representations of U.S. citizens relate to white-Americans’ physical traits, consumerism, economic, and political power whereas the French -save for their personal hygiene-draw a much more positive image in relation with gastronomy, fashion and clothes. Implications of these findings point at the role of mass media in the construction of stereotypes and attitudes and the need to foreground intercultural education in curricular programs.


Introduction
In a globalized world, intercultural encounters are increasingly common and become part of daily life.Colombia, a Latin American country, is becoming more globalized, entailing a need for intercultural studies to yield understanding on the relationships between Colombians and foreigners.As pointed out by Alred, Byram and Fleming (2002) geographical mobility has challenged the idea of homogenous societies and "[w]hat appeared to be mono-ethnic, -cultural andlingual societies are now complex multi-ethnic, cultural and -lingual polities.After the peace agreements of 2016, foreign investment is on the rise, tourism in Colombia has recently exploded (growth of more than 300% since 2006, Colombia Reports 2018), and numerous expats come to live in Colombia's main cities.This means that Colombians are now increasingly in contact with foreigners.In this backdrop, the study of social and cultural representations emerging from intercultural experiences gains relevance if goals of intercultural education are pursued.
This study draws from the interest of intercultural studies in exploring the social and cultural mindset of a social group (Colombians), in relation with other national cultures.U.S. and French nationals were chosen because of their proximity with the local collective psyche.U.S. citizens are the most familiar foreigners to Colombians due to their touristic, economic, and cultural presence.They constitute the largest group of visitors to Colombia -19,2% of the total international visitors-(Colombia Reports 2018).The U.S. has also long been Colombia's first commercial partner, with its omnipresent products on the local market and constant featuring of information related to the U.S. on the local news.Also, countless families have relatives living in the U.S. All these factors have led Colombians to construct complex representations of U.S. nationals.Moreover, the presence of North Americans and their culture has been heightened by the status of English language in the world (Alvarez Valencia, 2016).In particular, English has found a central position in the curriculum of Colombia to the extent that the government launched a national bilingual program in 2004 (MEN 2016).
In order to be able to compare and contrast the data related to U.S. nationals with similar data, we chose to examine representations about French nationals.Given their important role in Latin America and Colombia in the past, and -as reflected in Colombian media-as an important actor in the current global scene, French citizens and their culture are also familiar to most Colombians.French is the second foreign language most taught in Colombia, and historically it has played an important role in Colombia's educational system (González 2010).
The research aligns with the social constructivist perspective in intercultural communication that suggests that social representations (SRs) "constitute our realities" (Howarth 2002:8).Furthermore, gathering and analyzing such constructs enables researchers to observe the processes and forms of social thinking and reality appropriation of a social group.Nevertheless, despite the potential of social representations (SRs) to delve into intergroup relationships, the concept has been incorporated scarcely in intercultural communication research (Byram 2003;Carignan, Sanders & Pourdavood 2005;Dervin 2011;Hinton 2013;Holliday, Hyde & Cullman 2010;Kudo 2016).Social Representation Theory (Moscovici 1976(Moscovici , 1988) offers a suitable framework to explore the conception and perception that a group has constructed of another group and how this intergroup relation configures itself.Given that the prime function of SRs is to appropriate the unknown, that is, "to familiarize us with the strange, according to the categories of our culture" (Moscovici 1988:211), and that foreigners, by definition, are alien figures in relation to whom a variety of representations are constructed, the study of SRs comes as a particularly valid entry into intergroup relations.
Social Representations can be envisioned as a matrix of dynamic behaviours (Abric & Doise 2015).They constitute the mental substrate from which actions and interactional strategies grow.In this line of reasoning, this study aims to examine Colombians' social and cultural representations of U.S. and French nationals as a way to enquire into Colombian's perceptions, images and attitudes toward members of these two countries.This study contributes to the efforts of Colombia's national government concerning the need to increase cultural understanding and tolerance to differences.It provides insights into the nature of individuals' attitudes (hostile or welcoming) towards the outer groups mentioned.Results of this research will provide stakeholders with information to evaluate and plan educative intercultural campaigns to diminish and prevent possible future conflicts and help foster harmonious intercultural relationships between Colombians and foreigners.

Social Representations
Moscovici defines social representations as "a system of values, ideas and practices" (1973 xiii) that serves to establish a social order.It enables individuals to master their material and social world and to communicate with members of a community through a shared code for social exchange.Other authors define SRs as particular presentations of experiences, people, voices which are reinterpreted and represented or as an organized combination of cognitions relative to an object shared by the members of a homogeneous population in relation with that object (Flament 2001).Duveen (2001) emphasizes that social representations bind individuals together, permitting them to share knowledge, common practice and affiliations, thus supporting inclusion and exclusion systems.
A great deal of research on SRs has been produced since its emergence, touching upon a variety of contexts and domains such as education (e.g., teachers, administrators, students, public policy, curriculum, instructional systems or approaches, language teaching - Carignan, Sanders & Pourdavood 2005;De la Maza 2008;Maas, Miranda, Solís & Echeverría 2015); and representation of members of foreign groups (Amireault & Lussier 2008;Billiet, Maddens & Beerten 2003;Chryssochoou 2009;Hinton 2013;Wagner, Holtz & Kashima 2009).Within this last area of inquiry, several studies have explored stereotypes and attitudes towards either the French or North Americans.Generally, while in the US stereotypes toward the French are mostly negative (Rosenthal 1999;Drewelow 2011), people from other countries associate the European nation with fashion, cuisine and culture (Pugibet 1983, Verdaguer 1996;Henley 2011, Drewelow 2011;Pugibet 1983).Yet, stereotypes and attitudes toward North Americans are more nuanced, evoking admiration and criticism in relation to its historical and current socio political, economic and cultural influence in the world (e.g.Capozza, Trifiletti, Vezzali & Andrighetto 2009;Dodson 2005;Ladegaard 1998;Tan & Suarchavarat 1988;Willnat, He & Xiaoming 1997).

Sociocultural Representations
Social representations are directly related to culture.In fact, cultural theorists such as Stuart Hall used the term cultural representations, within his own framework of representations, conceived as the "production of meaning through language" (1997 16), while Sperber (1996) defined them as the "fuzzy subset of the set of mental and public representations inhabiting a given social group" (33).Duveen (2007), a leading figure in the field of SRs, explains the relation between the two concepts indicating that, Culture, then, can be taken as referring to a broader network of representations held together as an organized whole by a community.Social representations, in this sense, can be seen as particular cultural forms, and the analysis of social representations will always refer back in some way to the cultural context in which they take shape (545).
Drawing on Duveen's perspective, we use the concept of sociocultural representation (SCR) to express the symbiotic relationship between the social and the cultural dimensions.SCR determine how groups relate to each other and constitute therefore a central concept in the different fields that address intergroup relations, defined as "the way in which people in groups perceive, think about, feel about, and act toward people in other groups" (Hogg 2006:479).Representations play a key role in shaping the image that a national group constructs of a given foreign population and may have a direct impact on people's experiences of intercultural communication.Although in the field of intercultural communication few studies draw specifically on social representation theory (e.g.Hinton 2013), some interculturalists have started to discuss the concept of SRs in their work (Holliday, Hyde & Kullman 2010;Dervin 2011).

Sociocultural representations, attitudes, stereotypes and prejudice
The relationship between attitudes, prejudice and SCR is conceived in different terms.Rosenberg & Hovland (1960) define an attitude as the evaluation of an object according to three dimensions: affective (favorable or negative, sparked by a stimulus), cognitive (perceptions and beliefs), conative (disposition for action, oriented towards future or past).Attitudes are cognitive attributes, manifestations of personal and individual stance while social representations are social; they reflect group positions (Sammut 2015).We concur with Eagly & Chaiken (1993) in defining attitudes as the psychological evaluation of elements of an individual's environment according to a certain degree of favor or disfavor.Attitudes form part of representations and they interact dialogically at the collective and individual level, sometimes materializing in the form of stereotypes or prejudice.Stereotypes have been defined as the "social classification of particular groups and people as often highly simplified and generalized signs, which implicitly or explicitly represent a set of values, judgements and assumptions concerning their behavior, characteristics or history" (O'Sullivan, Hartley, Saunders, Montgomery & Fiske 1994:299-300).These authors argue that stereotyping is a "fundamental cognitive process of categorization" that requires simplification to help people make sense of the world (events, objects and experience) and organize common sense discourse (300).Stereotypes can, however, give rise to prejudice, a more undesirable cognitive and social predisposition to intergroup relations (McCage 2009).

Methods
Research on sociocultural representations of North American and French citizens in Colombia from the perspective of intercultural communication is inexistent to our knowledge, in part, because the area of intercultural communication is of relatively recent interest in this Latin American country (Álvarez Valencia, 2014).In view of this, the study draws on principles of qualitative-exploratory research that aims to investigate and build understanding of groups, processes, activities, or situations that have received little or no systematic examination (Marshall & Rossman 2011;Stebbins 2001).Marshall and Rossman (2011) highlight the relevance of this type of research because it helps to "identify or discover important categories of meaning and generate hypotheses for further research¨(p.69).One underlying rationale for this study is our interest to open the doors to more inquiries of this type, providing insight into the current work on intercultural communication, particularly, in the field of foreign language teaching in Colombia (Álvarez Valencia, 2014;Fandiño 2014).

Context and population
Due to practical constraints, this study was conducted in Medellín, one major city in Colombia that often attracts a considerable number of tourists.Medellín is the third largest city in the country with a population of around 2.5 million inhabitants.Participants of the study were a group of 200 (116 females and 84 males) individuals from all social strata whose ages ranged between 17 and 60.

Data collection
The research adapted a "free association" methodology ("association libre", also called evocation), designed by social psychologists (Abric 2003;Doise, Clémence & Lorienzi-Cioldi 1992).In this methodology, participants are invited to produce associations based on a theme-word/item provided by researchers.Several researchers (e.g.Vergès 1992;Wagner, Valencia & Elejabarrieta 1996) validate this type of research strategy indicating that free association is a cognitive dispositive in which words are more immediately called to express a representation.In the context of this study, a group of research assistants went to neighborhoods of different socioeconomic status of Medellín with the instruction to invite passers-by to take part in the study.Participants had to complete a questionnaire in which they were solicited to write down the first ten words that came to their minds when thinking about French individuals, then ten other words they associated with U.S. nationals.A total of 244 surveys were administered; 44 were excluded since participants provided incomplete information.Data analysis [1] The questionnaire responses were digitized in order to arrange and systematize data items based on variables such age, sex, profession, and education.The sociocultural representations of U.S. and French citizens were first analyzed through semantic analysis and the categorization of semantic fields.The purpose of semantic field analysis is to group all lexical units (words or noun phrases) that belong to a common conceptual domain and find the relationships between them and a general term (Kittay & Lehrer 1992).We used the UCREL Semantic Analysis System (USAS Semantic Tagset), a framework developed at the University of Lancaster and designed to conduct semantic analysis of text.The UCREL semantic tagger labels words according to 21 predefined discourse fields (e.g., "food and farming," "the body and the individual").By tagging the words semantically, we were able to identify salient themes in the data. [2] The analysis then focused on attitudes underlying each lexical item in the data set.We categorized words and noun phrases as carrying either a positive, neutral or negative semantic connotation.For example, "unpleasant", "greedy", "fat" were categorized as negative while "intelligent" and "hard-working" as positive.Such a process faces interpretative limitations since none but the person who mentioned the words knows whether they carry a positive or negative value.[3] Therefore, the analysis drew on the most common definitions and connotations of the words or phrases in the data .To ensure that the coding scheme was consistent and reliable, the researchers asked an external rater to conduct a third analysis.The inter-rater examination showed a high consistency rate between the three coding processes.

Findings
This section presents descriptive statistics of the analysis of the questionnaires.We focus on the sociocultural representations of U.S. and French citizens in terms of stereotypes and attitudes.We present the highest frequencies of lexical items for each country and then turn to analyze the categories under which all the lexical items were aggregated.Correlations between gender, education and attitudes are established.

Most common lexical items representing North American and French individuals
In total the participants produced 1999 items to characterize North Americans and 1998 for the French.Figure 1 displays statistics of the items that obtained more than 15 appearances in the data.As can be observed, more than 21 words exceeded the established cap, being "Tall" the most frequent word."Fair-skinned", "English" and "Gringo" followed, establishing a tendency that is recurrent in the data concerning the emphasis on phenotypical features of U.S. citizens.

Figure 1: Highest frequencies of lexical items representing North American citizens
With regard to the list of items associated with French citizens; with 64 appearances "The Eiffel Tower" is the referent that seems to be most commonly associated with the French culture.It is followed by words such as "Perfume", "Love", "Bread" and "Fashion."In accordance with general stereotypes about the French (Verdaguer, 1996), Respondents associated the French with gastronomy.Thus "Food and farming" fair as the most relevant category with 174 occurrences.Another significant representation (115 occurrences), materialized through the category of "Relationship: intimate/sexual," relates to another widely spread cliché that stereotypes the French as inclined towards romance, seduction, and romanticism.The semantic analysis also shows that the respondents focus both on the French's psychological traits (Personality traits: 100) and their physical appearance "Anatomy and physiology" with 80 occurrences."Clothes and personal belongings" with 86 repetitions converges again with the commonly circulating representations of France as the country of fashion (Verdaguer, 1996).In the same vein, French culture is associated with art ("Arts & Crafts": 117 occurrences), sports ("Sports and games generally": 46), and education (50).These representations can be rooted in Colombian media that often provides content referring to French artists, cinema and education, and important sporting events such as the Tour de France.The analysis also indicates that French culture is familiar to the respondents because of a variety of icons, including "Geographical names": Paris; and tourist spots: the Eiffel Tower.
In South and North America, a widespread prejudice circulates that the French have poor personal hygiene based on the idea that they seldom wash themselves (Henley 2011).This prejudice is echoed here within the category of "Cleaning and personal care" with 55 occurrences.
Figure 4 presents the results of the semantic tagset analysis concerning U.S. nationals.The category with the most frequencies is "Geographical names", followed by "Anatomy and physiology" with 194 occurrences.This classification, along with the category "Judgement of appearance (pretty etc.)" ( 20) indicates that Colombians are particularly sensitive to the phenotype of Anglo American individuals.The psychological dimension is also of interest to Colombians who foreground categories of "Personality traits" (70), "Approachability and friendliness" (34), and "Ability: Ability, intelligence" (59 items).

Figure 4: Most recurrent USAS semantic categories associated with U.S. citizens
The semantic analysis sheds some light on participants' perception of U.S. nationals.Americans are seen as rich ("Money: Affluence": 36), powerful ("Power relationship": 88), belligerent and destructive ("Warfare, defence and the army: Weapons": 33) and "Damaging and destroying" (27).These perceptions can be related to the influence of U.S. foreign policy in South America and worldwide.
With a frequency of 83 items, "Food" connects semantically to "Health and disease" (92).The former consists essentially of words related to junk food while the latter involves words related to overeating and obesity.Other categories such as "Sports" (40), "Music and related activities" (30), "Drama, the theatre & show business" (20) are greatly rooted in the local media and entertainment industry where their cultural products occupy a large market share.Colombia is a major consumer of movies and T.V. shows produced in the U.S.. Several U.S. channels (CNC, ESPN, Discovery) have Spanishspeaking divisions and major sporting events (NBA, boxing, NFL) are broadcasted in Colombian channels.Finally, unlike the semantic tagset results from the list of French words, the tagset about U.S. nationals features English language ("Language, speech and grammar": 42) as a relevant marker associated with North Americans.

Attitude analysis
Sociocultural representations, materialized in the shape of stereotypes and/or prejudice, are necessarily underscored by attitudes.Attitudes comprise neutral, negative or positive subjective judgements of entities.Figure 5 compares the language produced by the participants in terms of neutral, negative and positive attitudes towards North American and French individuals.What is interesting in this data is that negative representations associated with U.S. nationals (508 items= 25.4%) predominate over their French counterpart (215 items=10.8%).Likewise, U.S. nationals rate lower as regards positive representations (418 items=20.9%),while the French seem to evoke more positive images with 673 (33.7%) occurrences of such items.

Gender variable
It is also relevant to establish whether women or men hold more positive or negative attitudes toward citizens of the nations involved in the study (see figure 6).Overall, female participants produced more connotative language than male participants.Results show that females' image of French people is more positive (36.2%) than that of male participants (30.2%) while females' image of U.S. citizens is consistently more negative (25.7%).

Figure 6: Comparison of male vs. female attitudes
We also analyzed the relationship between attitudes and gender by country.We examined all items produced by men and women that expressed positive or negative valuation about French and North American individuals.Positive attitudes towards French individuals predominate in both genders; women with 36.2% and men with 30.2%.Concerning negative attitudes women display a tad more negative view (11.5%) towards French than men do (9.8%).By contrast, U.S. nationals are represented in less positive ways in the data although differences between females and males are narrow: positive (21.6% vs. 20%) and negative attitudes (25.7% vs. 25%).As can be seen, the survey participants hold more positive sociocultural representations about French individuals than North Americans, with females producing more connotational words regarding the French.

Attitudes and level of education
Concerning the relationship between level of education and attitudes, Figure 7 shows that for the French neutral lexical items obtained the highest percentages, being college and technical schools students the least judgemental population with 60.5% and 57.7%.Likewise, college students use words or phrases that depict a more positive attitude (e.g.splendid, purity, urban beauty) towards French people with 33.8%, compared to 8.5% that featured items evoking negative connotations (boring, anorexia, infrequent shower).Participants with elementary school education and postgraduate education reported significantly more connotative items with 17.8% (e.g.unkempt, impudent) and 38.9% (e.g.well dressed, clever) for the former, and 16% (e.g.presumptuous, they do not bathe) and 39% (e.g.faithful, educated, beautiful lights) for the latter.Although one could expect to find more stereotypes and prejudicial attitudes among lower educated segments of the population and a reduction of them among more educated groups, it appears surprising that participants with graduate education rate higher than anybody else.These data must be interpreted with caution though.One of the reasons for these two educational levels to stand out is the sample size (see Table 1): only 9 participants belonged to the lowest education group (elementary school), while only 10 participants reported to have postgraduate education.With regard to the respondents' representations of North Americans, it is apparent from the data in Figure 8 that prejudicial attitudes against U.S. nationals are consistently more significant than the French.Through utterances such as racism, exploitative, obese, arriviste, the participants within the categories of elementary school and postgraduate level report the highest percentage of negative items, with 26.7% and 37%, respectively.The educational levels of high school and college rate higher in the number of negative expressions to represent U.S. individuals and their culture (e.g.solitary, not expressive, consumerist, junk food) compared to the data analyzed about the French above.Apart from the category of postgraduate level, the statistical results for each educational level (Elementary, High School, Technical and College) do not differ quantitatively much from each other, which could sustain the thesis that stereotypes or prejudicial attitudes (either positive or negative) about the North American culture are more widely shared and cemented in the Colombia's mindset.

Discussion
In this study respondents were asked to list the words that came to their minds when they thought of French and North Americans individuals.The participants focused on aspects such as physical appearance, tourist attractions, personality, food customs, and personal hygiene.Words such as "Eiffel Tower," "Perfume," "Love," "Bread," "Perfume" and "Art" score higher for the French, while "Tall," "Blonde," "English," "Gringo, "Dollar," and "War" received the highest ratings for North Americans.The semantic analysis that aggregated all of the items produced for each country in different fields resonated well with the list of most frequent words, in particular with the representation of U.S. citizens and the most rated category: Anatomy and physiology.The focus on phenotypic features of Americans and food in the case of the French, two of the aspects that most distinguish both cultural groups from the Colombian culture, is in part explained by what Wagner, Holtz, and Kashima (2009) point at that, "In the majority of cases the attributes defining a social group are not accidental but the result of an attempt to distinguish members of the ingroup from others" (p.369).In this regard, it makes sense that the ingroup, Colombians, highlight physical features such as height and hair color from Americans, traits that are not common in the Colombian phenotype.Although physical attributes also appear in the list for the French, they do not appear as central, perhaps because physically-wise they seem to be perceived as more similar to the ingroup.The ingroup's differentiation process or "Othering" constitutes a device used to affirm and reinforce their own identity by constructing sameness and difference (Abu-Lughod 1991; Holliday 2011).Othering is a process of objectification and "creation of the other" who is lumped together into a seemingly homogeneous social group, disregarding the subjectivity and complexity of the individual.Sociocultural representations in the shape of stereotypes not only contribute to "other" members of the outgroup but also produce essentialized views of them.Although Wagner, Holtz, and Kashima (2009) indicate that judgements about ingroup or outgroup members informed by essentialist thinking can easily switch to discrimination and racism, they also add that "essentializing is a versatil representational tool [...] that is used to create identity in groups with chosen membership in order to make the group appear as a unitary" (p.373).Colombians, however, are not alone in their sociocultural representations of the members of these two outgroups; a review of previous research shows similar stereotypes of French and North American citizens.
Compared to other studies about stereotypes concerning the French, the results coincide with widespread representations of members of this outgroup (Drewelow 2011;Pugibet 1983;Rosenthal 1999), although it is noticeable that in this study most negative representations reach lower scores.A look to Figure 2 shows that all lexical items (e.g.Eiffel Tower, Perfume, Fashion, Art, Romantic, Wine, Gastronomy) do not imply negative stereotypes.The list of items evokes Pugibet's (1983) study with 151 Mexican college students who participated in a task of association in which they wrote a list of words prompted by three lexical items: France, the French and the French language.The author reports that the most common themes included: wine, the Eiffel Tower, physical features, gastronomy, perfumes, and fashion among others.
With respect to negative stereotypes, authors like Rosenthal (1999) and Drewelow (2011) in the North American context state that traditionally the French have been consideredrude, arrogant, uncooperative, unfriendly, unhygienic, disdainful, licentious, weak and coward.Although some of these words are mentioned by the participants in our study, personal hygiene being very frequent, most of the words do not rate strongly as significant representations of the French.From comparing stereotypes attached to the French in countries such as Mexico, U.S. and Colombia, one can deduce that selfrepresentation of the ingroup and historical relationship with the outgroup contribute to the creation and maintenance of stereotypes.One example of this is North American's sociocultural representation of the French as weak and coward, a stereotype that is not present in Colombian's rep resentation of the French.According to Rosenthal (1999) these stereotypes trace back to the long historical tradition of rivalry between the two nations.The author adds that during colonial times "The hereditary enemy of Britain became the natural adversary of the British colonists in America, who absorbed, adapted, and perpetuated the jaundiced views of the mother country" (p.905).
While stereotype and prejudice may be rooted in historical events, they also emerge from self-perception of the ingroup.A case in point is Colombians' view of the French and North Americans as cold (and associated words such as coldness), a perception that American and French individuals do not hold about each other.Colombians' perceive themselves and are said to be warm and happy people which, in part, explains why they find the two outgroups under discussion to be cold.In turn, American and French individuals may not find this trait as a differential feature that distinguishes both cultural groups.Thus, governing stereotypes assigned to certain cultural groups crystallize as a result of differentiation processes or historical inheritance.
Overall stereotypes about French nationals seem to be more stable across different countries, while for North Americans the results are less comparable to other studies that have explored the subject in other contexts (Capozza, Trifiletti, Vezzali & Andrighetto 2009;Dodson 2005;Glick et al. 2006;Ladegaard 1998;Stephan et al. 1993;Tan & Suarchavarat 1988;Willnat, He & Xiaoming 1997).In countries such as Thailand, U.S. citizens are constructed as pleasure loving, athletic, individualistic, sensual, efficient, musical, scientifically minded, artistic among others (Tan & Suarchavarat 1988).By contrasts, a more recent study examining the relationship between media exposure and stereotypical perceptions and feelings toward North Americans in Hong Kong, Shenzhen (China) and Singapore concluded that the most common stereotype among 625 college students about Americans is that they are "open-minded," "individualistic," and "aggressive" (Willnat, He and Xiaoming 1997).Russian individual perceived North Americans to be "adaptable, proud, optimistic, patriotic, independent, outgoing, competitive, aggressive, progressive, materialistic, emotional, friendly, egotistical, impulsive, and flexible" (Stephan et al. 1993, p. 57), while Italians position them as allies, imperialist and colonialist (Capozza et al. 2009).
As can be seen, stereotypes about Americans vary from context to context although some images repeat across countries, particularly, visions of the U.S. and its citizens as powerful, individualist, aggressive and materialistic.In reviewing the results of the semantic fields and the frequency of words associated to U.S. nationals in our data, several words related to these stereotypes standout, to wit: "war" (23 times), "money" (15), "dollars" (23), "capitalism" (11), "power" (10), "capitalist" (7), "wealthy" (7), "consumerism" "individualistic," (3), and "aggressive" (2).Nonetheless, as presented in Figure 1, Colombians' representations focus more on "Anatomy and physiology," "Health and disease," "Food," Personality traits" and "Money-generally," being this last semantic group the one that connects the most with representations of the outgroup members as materialistic.
Besides political, economic and nationalistic factors, gender and educational categories also take part in constructing sociocultural representations.The findings in this study differ from previous studies where female respondents displayed less prejudice than their male counterparts (Navarrete, McDonald, Molina & Sidanius 2010;Wierzbicka 1997).Women's representations are slightly more polarized than men, especially with regards to the positive views of the French, and a tad more critical in their attitude towards U.S. citizens.Likewise, our results differ from previous studies that showed a negative correlation between education and stereotype/prejudice (Guthrie, King & Palmer 2000;Ostapczuk, Musch & Moshagen 2009;Wagner 1995).For example, Radloff (2004) has provided evidence of the impact of higher education on racial prejudice.Formal education, by providing knowledge on minorities while teaching people to recognize and understand prejudice, increases individual's capacity to reject stereotyping and prejudice.In our study, there is no significant difference between the number of connotative words produced by respondents with higher education and those without it.Overall, this is an indicator that Colombian respondents hold a homogeneous representations and attitudes of U.S. and French citizens across educational levels.

Conclusions
The linguist Anna Wierzbicka (1997) once said that "Language-and in particular, vocabulary-is the best evidence of the reality of "culture," in the sense of a historically transmitted system of "conceptions" and "attitudes" (p.21).In agreement with this statement, this study drew on the participants' word associations to examine their sociocultural representations.Results evidence that to a great extent stereotypes and attitudes toward citizens of France and the U.S. are informed by social, cultural, political and economic dimensions projected by these countries and usually disseminated by media.However, stereotypes and attitudes also grow out of the ingroup's self-representation and identity features, which are evidenced when ingroup members engage in processes of differentiation or "othering".Generally, the participants of the study have more favorable stereotypes and attitudes towards French individuals than North Americans.They mostly associate the former with aspects of high culture while Caucasian physical traits, consumerism, economic and political power came to their mind when they thought of U.S. citizens.
In inquiring into what may be a probable source of the prejudices and stereotypes identified in this study, we agree with Van Dijk (2000) that the discourse of the media plays a prime role in mobilizing and creating sociocultural representations.In the case of Colombia, the principal source of information is television, with 72% of individuals following news about international affairs (Tickner, Cepeda, Rouvinski & Milanese 2016).Added to this is the particular preference of local channels for North American shows and movies.This abundance of exposure to foreign media and market goods in part explains Colombians representations of North Americans concerning their physical traits, values, and cultural practices.Such a conclusion evokes previous research that has pointed at the positive correlation between media and stereotype/prejudice construction towards foreign outgroups, including the U.S. and France (Ladegaard 1998;Pugivet 1983;Rosenthal 1999;Tan & Suarchavarat 1988;Wagner, Holtz & Kashima 2009;Willnat, He & Xiaoming 1997).
Despite their ubiquity, negative and positive stereotypes and prejudicial attitudes should be looked at critically and different awareness-raising actions should be taken from an educational standpoint.Different scholars have shown that intergroup antipathy, stereotypes and prejudice emerge or dissipate within short spaces of time (Hinner 2010;Hogg & Abrams 2006).This opens the doors for possible action in the different educational levels where intercultural communication should be part of the curriculum.Nonetheless, although intercultural education is necessary, as Allports' (1954) Contact Theory has expounded, encounters with outgroup members have shown to be the most effective way of calling into question stereotypes and reducing prejudice.
The negative and positive stereotypes and prejudicial attitudes of the participants of this study constitute a call for stakeholders to implement intercultural education programs in Colombia.Suárez's inventory (2009) of intercultural programs in Colombia provides evidence of such deficiency.So far most programs have focused on etno-education, multiculturalism and bilingual education, involving indigenous and Afro-Colombian communities; nevertheless, intercultural education programs should also tackle intercultural encounters between Colombians and members of other countries, such as the U.S. and France.

Figure 2 :
Figure 2: Highest frequencies of lexical items representing French citizensSemantic analysisFigure3displays the most recurrent categories (above 20 frequencies) associated with French culture that were identified using the USAS semantic tagset.

Figure 3 :
Figure 3: Most recurrent USAS semantic categories associated with French citizens.

Figure 7 :
Figure 7: Correlation between level of education and connotative items associated with French nationals.

Figure 8 :
Figure 8: Correlation between level of education and connotative items associated with U.S. nationals.

Table 1 :
Characterization of population