Role of Language in the Formation of Social Attitudes

2423 words (10 pages) Essay

23rd Sep 2019 Languages Reference this


Disclaimer: This work has been submitted by a university student. This is not an example of the work produced by our Essay Writing Service. You can view samples of our professional work here.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of

The Role of Language in the Formation of Social Attitudes

Does language affect our attitudes about others? Early theories on attitudes saw them as stable, unchanging mental representations that dictated consistent behavior (Ogunnaike et al., 2010). Recent research suggests that the context in which these attitudes are expressed has an effect on the attitudes themselves. The influencing context can range from thoughts or events preceding attitude expression, who it is being expressed to, or even by the weather. This dispels the notion of attitudes as static representations, and is evidence in favor of attitudes being influenced by many social and environmental forces. (Ogunnaike et al., 2010).

Results from studies about the impact of social and environmental contexts and forces on attitudes can provide evidence for a relativistic stance. If our thoughts are influenced by social and environmental features of the world around us, then the language in which these attitudes are expressed may also play a role, given that language features largely in the uniqueness of the social and cultural setting in which these attitudes are expressed. This harks back to Benjamin Lee Whorf’s hypothesis that language influences thoughts and judgments. If Whorf is correct, then assessments in different languages should engender different thoughts and responses.

In their 2010 study, Oludamini Ogunnaike, Yarrow Dunham, and Mahzarin R. Banaji studied the relative automatic or implicit attitudes of bilinguals toward the same two attitude objects while varying the language of assessment. They performed two experiments that both used the implicit-association test (IAT), which measures the automatic connections or relationships made between objects and attitudes.

The first experiment was conducted in Morocco, with the participants being native Moroccan bilinguals fluent in French and Arabic. The predicted result was expected to show that more favorable evaluations of names would occur if the names and words used in the experiment were in the test language.

Participants were asked to perform two categorization tasks, with one block having seven Moroccan names sharing a response key with positive Arabic words and another with the same seven names being associated with negative Arabic words. Another version of the test used the same construction but using French names and French positive and negative words. The language condition was the independent variable, and the IAT D-score was the dependent variable, with positive IAD D-scores indicating a preference for Moroccan over French names. Subtracting a participant’s IAT D-score of the French version from the D-score of their Arabic version gave what they referred to as the language effect score (Dlang) (Ogunnaike et al., 2010).

Because French and Arabic use different scripts and have auditory disparities, Ogunnaike and colleagues presented stimuli in written and aural forms, with half of the participants performing under the written condition and another half performing under the aural condition. Following the two IATs, participants self-reported their years of studying Arabic and French, degree of fluency, and attitudes towards Morocco and France.

The participants demonstrated a significant preference for Moroccan over French people when the test was administered in Arabic, but showed equal preference for Moroccan and French people when the test was administered in French. 80% of participants (32 out of 40) displayed more pro-Morocco attitudes when tested in the Arabic condition. In the self-reported data, participants reported a greater preference language-wise for Arabic over French, and for Moroccan over French people. A positive correlation between a participant’s self-reported preference for Arabic over French and Moroccan over French people and their Dlangscore would indicate that their self-reported preference is driving the language effect seen in the data, but no correlation showed significance, which suggested that the language effect does not stem from a global preference for a particular language or population, or from overall language ability. Associations within the languages seem to be the source of the language effect (Ogunnaike et al., 2010). 

To test the replicability of the first experiment, Ogunnaike and colleagues conducted a second study in the United States, with self-reported Spanish and English bilinguals as participants. A similar IAT design was used, but with English and Spanish instead of French and Arabic. Stimuli were again varied from written to auditory as a between-subjects factor. In this test, positive IAT D-scores would demonstrate preference for Hispanic over Anglo names.

The results of the second experiment demonstrated that when participants were tested in Spanish, they showed a significant preference for Hispanic names. When tested in English, they not only showed no preference, but had negative D-scores. The language effect score was calculated as in the first study, and indicated stronger pro-Hispanic attitudes in the Spanish-language condition, with 71% (27 out of 38) of participants displaying the predicted preference. In the self-report data, a greater preference language-wise for Spanish over English, and for Spanish-speaking people over English-speakers was found. No self-report responses were correlated with the language effect score, demonstrating that the effect of language on attitude was not related to a preference for a specific language or its speakers (Ogunnaike et al., 2010).

The overall results of the study are consistent with the researchers’ predictions and demonstrate that language may influence automatic attitudes, prompting a greater preference for the population associated with that language. The replicability of the study demonstrates that the methods used are satisfactory. However, some conditions in the study were not replicated exactly, which may have contributed to the negative IAT D-score seen in the second experiment.

The first experiment was conducted in Morocco with native Moroccans, who as the study describes as having been exposed to both French and Modern Standard Arabic in school, signage, and media. In contrast, the second experiment was conducted in the United States, with no indication of the Spanish-English bilingual participants being native to the United States. If a number of the participants were immigrants from primarily Spanish-speaking countries, then a much stronger preference for Spanish is to be expected given that they may have been exposed to primarily Spanish signage, media, and school instruction prior to immigrating to the United States. A study conducted in a country, with participants native to that country who were exposed to both Spanish and English in school, signage, and media would probably display results closer to the ones seen in the Moroccan experiment. In addition, the researchers attributed data loss in the Moroccan experiment to a lack of resources or practice with computerized testing. Conducting the test in a Spanish-speaking country where the population has somewhat similar technological capabilities and resources to Morocco may have also given the researchers different results.

I would also like to see if this study could be replicated in other languages as well, in countries with similar technological capabilities and resources as the United States, such as seeing if a study conducted in English and German contributes to a preference for Germans in Germany. The researchers themselves stress the importance of test language in conducting a similar study using the IAT to measure implicit attitudes. They also point out that the exact mechanism through which language influences thought must be further investigated.

The results of the study conducted by Ogunnaike and colleagues seem to provide support to Whorf’s relativism hypothesis. The language that participants were exposed to influenced their implicit attitudes. The study demonstrated that more than being just a conduit for communicating preference, language can be involved in building stereotypes and assumptions of a given population.

Another study that has gleaned results supporting Whorf’s relativism hypothesis is the study conducted by Shai Danziger and Robert Ward. Their study sought to determine if language could influence the accessibility of social associations (Danziger and Ward, 2010).

The Danziger and Ward study tested whether the implicit associations of the Arab Israeli participants differed based on the language used in testing (Hebrew or Arabic). Given the previous evidence that thoughts are context-dependent, it was predicted that participants would demonstrate more positive associations to Arabs and less positive Associations to Jews when Arabic was used as a test language.

The population used in the study were bilingual Arab Israeli undergraduates from Hebrew-speaking schools. The study outlined that for many Arab Israelis living in Israel, Arabic is spoken at home, with exposure to Hebrew starting in the third grade and continuing throughout secondary school, or through higher education if they are studying at a Hebrew-speaking university, where Hebrew is the language of instruction in all non-language courses.

Similar to the study conducted by Ogunnaike and his colleagues, the Danziger and Ward study used the IAT. Two versions of it were designed for this study. One was an Arab-Jew IAT that used Arab and Jewish names along with trait words with positive or negative connotations. The first block had Arab and positive traits sharing a response key, and Jewish names sharing a response key with positive traits. In the second block, the response mapping was reversed, with Jewish names sharing a response key with positive traits, and Arabic names sharing a response key with negative traits.

The second IAT used in the study was used to verify that language dominance did not drive the language effect seen in the results of the Arab-Jew IAT. In this IAT, weapons and instruments were associated with either positive or negative trait words. The blocks were arranged similarly as in the initial IAT. Their independent variable was the language used to conduct the study, and the dependent variable was the IAT D-score (Danziger and Ward, 2010).

Positive IAT D-scores indicated an implicit preference to Arabs over Jews, and musical instruments over weapons. D-scores were found to be higher for Arabic than Hebrew in the Arab-Jew IAT, and positive associations to Arab names and musical instruments were observed. These results are in line with the researchers’ prediction that the language used in the experiment influences the attitudes expressed.

The researchers did indicate that there may be a possible priming effect that was not measured or observed in the course of their study. One priming effect could be that the language used in testing activates memories that were experienced in that language, leading to association of attitudes to the said experiences and not necessarily just to the language. Another priming effect could be that language primes norms and traits characteristic to speakers of the language that facilitate healthy social interaction. Overall, they believe that their findings demonstrate that language use influences attitudes, and that bilinguals’ thoughts and attitudes about their social world varies depending on the contextual language (Danziger and Ward, 2010).

I find the methods used by Danziger and Ward to be satisfactory for the hypothesis they were pursuing. They consider possible confounds and give context to the linguistic background of their participants. I agree with their assessment that their results are in accordance with the relativistic perspective of language influencing, but not determining, thought and cognition.

Having taken the results of both studies into account, I am fairly confident in coming to a conclusion that the language used in expressing an attitude or preference has some effect or influence on the attitude itself. Beyond being a mere medium for the expression of thoughts and attitudes, the results of both studies demonstrate that it plays an integral role in the development of thoughts and feelings regarding our social environment.

The similarity of methods and results in both of the studies discussed above indicate that the experimental design used is replicable. For the study conducted by Ogunnaike and colleagues, I have no issues with their design for the initial experiment, just some reservations about how they chose to replicate it. Both studies used Arabic as one of the languages of concern. Since the population used in the Danziger and Ward study seem to have possibly strong cultural ties to Arabic (as it was indicated to be their primary language prior to being exposed to Hebrew in their education), I would be interested to see this study replicated in another country or culture (such as Western European or East Asian), as the strong language effect demonstrated in both studies could be attributable to a non-observed component unique to the Arabic language. I would like to see if the language effect produced in both studies is replicable across many other cultures and languages. I would also be interested in finding out whether or not the language effect observed is generalizable. For example, if having participants read or listen to a news report detailing a heroic gesture of a Moroccan national in both Arabic and French increases pro-Moroccan sentiments.


  • Danziger S., & Ward, R. (2010). Language changes implicit associations between ethnic groups and evaluation in bilinguals. Psychological Science, 21, 799-800
  • Ogunnaike, O., Dunham, Y., & Banaji., M.R. (2010). The language of implicit preferences. Journal of Experimental Social Psychology, 46 (2010), 999-1003

Cite This Work

To export a reference to this article please select a referencing stye below:

Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.

Related Services

View all

DMCA / Removal Request

If you are the original writer of this essay and no longer wish to have your work published on the website then please:

Related Lectures

Study for free with our range of university lectures!