A Cross-Cultural Comparison of Letters of Recommendation Essay Sample
- Word count: 4883
- Category: cultural
A limited time offer!
Get a custom sample essay written according to your requirements urgent 3h delivery guaranteedOrder Now
A Cross-Cultural Comparison of Letters of Recommendation Essay Sample
Abstract–Letters of recommendation (LRs) from different countries are as individual as the local academic cultures from which they arise. Distinct regional patterns emerged in this comparative study of letters of recommendation from the United States, United Kingdom, Germany, and Eastern Europe. Two types of analysis were performed: first, a quantitative analysis examined features such as linearity, symmetry, data integration, advance organizers and sentence types; second, a qualitative analysis examined the content of the sections of the letters.
Differences were found cross-culturally in the quantitative analysis. Significant differences were also found in the organizational patterns and methods of support. Organizational patterns varied from topical to chronological organization. LR writers from different regions supported their recommendation of the applicant with different types of evidence, from factual lists of achievements to storytelling. The format of the letters themselves showed similarities cross-culturally. 1998 The American University. Published by Elsevier Science Ltd. All rights reserved Introduction A great deal has been written in the past decade on cross-cultural differences
in academic writing; not surprisingly most of this attention has been focused on the research article. Little notice has been given to the less public texts of the academic community, aptly named occluded genres in Swales (1996). The purpose of these texts is primarily to conduct the business of the academic community–requesting reprints, recommending students, reviewing articles, evaluating colleagues, and so forth. Since these occluded genres are private documents, they are much more likely to retain their authors’ cultural influences than are the more public, highly stylized
texts such as research articles. That is, according to Swales, the rhetorical patterns of the L1 would likely be more pronounced in these private, occluded texts than in research articles, and therefore the study of the occluded genres might yield some interesting cross-cultural comparisons. Cross-cultural comparisons of texts promise to give insight into the differences in the way that the same intention is expressed from one culture Address correspondence to: Kristen Precht, Department of English, Northern Arizona University, Box 6032. Flagstaff, AZ 86001, USA. 241 242 K. Precht to another.
Though intentions may be the same cross-culturally, differences in verbal and nonverbal expression are as subtle as the degree of directness that’s considered appropriate in making a request, or as obvious as a bow vs a handshake. A comparison of ways cultures express themselves in writing is undertaken in contrastive rhetoric. In examining the discourse of letters themselves, one can see elements of both spoken and written discourse, as well as a great deal of variety, from the very formal business letter to the chatty holiday newsletter, each type with a host of expectations as to form, structure, and content.
Within academia itself there are many types of letters, from submission letters, to reprint requests, to correspondence with colleagues and editors. Amidst all the academic correspondence, letters of recommendation for graduate school are a particularly interesting text type to study cross-culturally; they are occluded, and representative of local academic culture. LRs themselves can vary to a certain degree within a given cultural context (recommending either admission, or funding, or advancement), but the intention of recommending a student or junior colleague allows for a great deal of functional similarity in LRs, as most admission or fellowship committee members will readily attest.
Bouton (1995) highlighted important differences in the structure of LRs from different Asian countries and the United States; the present study seeks to look more carefully at cultural differences in discourse features, content, and semantics. The cross-cultural study of LRs also highlights a more practical aspect of contrastive rhetoric research. As the graduate programs in the United States become more and more international, LRs are exchanged between academics all over the world. No longer can LRs be interpreted through the lens of one’s local academic culture.
Though LRs all seem to have the same purpose, the ways that writers from different cultures express support vary a great deal. This variation can have a profound effect on a student’s admission prospects for graduate programs. This study was conducted to compare the structure and support in LRs from four regions: United States, United Kingdom, Germany, and Eastern Europe. A very narrowly focused corpus of LRs was identified in order to examine the generic structure of the letters and the preferred ways to support the writer’s recommendation.
While this study appears to be among the first studies of LRs, research on other aspects of academic discourse has been voluminous, from reprint requests (Swales 1990) to politeness strategies in scientific articles (Myers 1989). One function of academic discourse is to serve as gatekeeper by requiring the ability to use academic discourse appropriately for inclusion in academic ranks (Cooper 1989; Purves & Hawisher 1990; Berkenkotter & Huckin 1995). Another function of academic discourse is to serve as a social medium with established norms for the network of like-minded people in the academic community (Blanton 1994).
Writing a LR involves both the gatekeeping and social functions of academic discourse: The task of the writer is to evaluate the qualifications and potential of applicants for memA Cross-cultural Comparison of Letters of Recommendation 2/45 bership in her own community. As with other academic writing tasks, the LR is written by one academic for other academics to read, but unlike other academic writing tasks, there is no specific training in how to write a LR. Any similarities within a local culture or region are then due to the influence of local discourse conventions.
As such, the LR is a good example of the rhetorical structures, patterns, and formulas for a region. One of the primary purposes of the LR is to share one’s evaluation of the applicant, and so we are concerned with identifying the ways evaluation is expressed in different cultures. Some important variables have been noted in Clyne (1987, 1991, 1994) between German and English texts, with additional mention made of differences between British and American English. Clyne noted cultural differences in academic texts, notably that a cluster of traits correlated in the “weU-organized” American or British text: linearity, early advance organizers, data integration, and symmetry (1991, 1994).
However, in the German texts, correlations between these features were not significant. Furthermore, Clyne determined that German academics tended to follow German textual patterns when writing in English with respect to features he considered to be strongly based on sociocultural patterns, such as linearity and symmetry. In factors that are less deeply ingrained socioculturally, such as data integration, sentence types and advance organizers, German academics seemed to follow English patterns (1991).
From Clyne’s comparative work, we would expect that the English and American texts would tend to be linear, symmetrical, have data integration and have early advance organizers; we might expect the German texts to have evidence of data integration, advance organizers, and topic sentences, with less evidence of linearity and symmetry. As for Eastern European texts, we know very little about culturally preferred discourse patterns. In one case, Connor (1996) argues that Hungarian texts tend to use narration where an American or Western European text would be factual or descriptive. With respect to Eastern European or Slavic
academic texts, there appears to be very little research, and so the present study of LRs provides a first exploration. In the case of the present study, since the LR is ultimately read and interpreted by Americans, it’s important to consider how the letter would likely affect an American reader. This effect is not only formed at the macrolevel by differences in structure, politeness, directness and so forth, but also in slight differences in phraseology (Swales 1996; Maier 1992).
The differences in the formulas for recommending, though seemingly slight, will also therefore be considered in this cross-cultural comparison of LRs. Method Corpus Thirty-nine LRs were identified, 10 each from the United States, United Kingdom, and Eastern Europe, and 9 from Germany. The LRs were all 244 K. Precht written for applicants to the same graduate program in law, and they were chosen randomly from the files of applicants who had been admitted to this program. The caliber of the applicants cross-regionally was determined to be similar by admissions officers who evaluated applicants’ grades, the caliber and reputation of their educational institutions, the statements written by the applicants, and TOEFL scores.
Permission to use these LRs was granted on the condition that confidentiality be respected, and that all identifying features including names of applicants, faculty, universities, nationalities, and in certain cases identifying characteristics be changed or deleted. Though the corpus for this study is small, the exceptionally narrow focus of this group of letters–written by academics in the same field, for applicants of similar caliber, to a single degree program–creates a specialized corpus (Renouf 1987), appropriate for contrastive rhetoric research.
Developing such a narrow focus in a corpus improves one’s ability to accurately identify and contrast the rhetorical patterns in comparable texts of different cultures. All LRs were written by academics or professionals in law except for one (E14-EE, which was written by a businessman). The American and British LRs were written by native speakers (NS) of English, as far as could be determined, and the rest by non-native speakers (NNS) of English. For identification purposes, each LR has a code identifying first the applicant by letter (A-V), second the author of the letter (1-60), and third the country or region that the letter is from (UK for United Kingdom, EE for Eastern Europe, G for Germany, and US for United States)*.
Unfortunately it was not possible to secure a larger number of letters from any one language group or country in Eastern Europe, and so a compilation of 10 letters from four countries was assembled. This compilation is problematic in that the rhetoric and academic discourse conventions between these language groups is collapsed. The results from the Eastern European region, therefore, should be treated very tentatively. However, they do tend to exhibit some similarities, and make an interesting comparison to the American, British and German patterns.
The composition of the Eastern European writers is as follows: the LRs from Russia were E12-EE, E13-EE, and E14-EE, the LRs from the Czech Republic were P46-EE and P4 7-EE, the LR from Hungary was B4-EE, and the LR from Poland was K29-EE. The remaining 3 LRs were from an international Eastern European academy, and therefore not identified by nationality. Of the 39 LRs, 21 were written for male applicants, 18 for female. By region, the German letters were all written for females; Eastern Europe had 6 for males, 4 for females; the United Kingdom had 8 for males, 2 for females; and United States had 7 for males, 3 for females. Nearly all LRs were written by males.
*Thus the LRs B4-EE and B5-US are written for the same applicant (B) by two different professors (4 and 5), one from Eastern Europe (EE), and one from the United States (US). A Cross-cultural Comparison of Letters of Recommendation TABLE 1 Length of Letters (in number of words) 245 American British German Eastern European Average 454 303 294 350 Median 302 303 295 312 Range 1006 253 276 509 The letters ranged in length from 98 words to 1186 words. Although the American letters had the highest average length, the British median length was actually one word longer than the American median (see Table 1).
The difference in the range of lengths within a region (the shortest text’s length subtracted from the longest) is clearly responsible for the higher average length in American letters, with the British range being about a quarter the size of the American range. The German median length was shorter than the British and American medians, and the Eastern European letters slightly longer. Procedures The first haft of the analysis focused on comparing textual features cross culturally. In determining how to focus the comparative analysis, different models for contrastive rhetoric studies were consulted.
Clyne (1987, 1991) identifies several areas of divergence between American and German texts which were applicable to this corpus of LRs: digressiveness/linearity, textual symmetry, data integration, advance organizers and sentence types. These elements figure prominently in other analytic models for contrastive analysis (Purves & Hawisher 1990; Kaplan 1995) (see Fig. 1).
Clyne’s (Clyne 1991) framework and Clyne’s (Clyne 1987) definitions of characteristics have been adopted, with some modification, for the present study. Ciyne, 1987, 1991, 1994 Purves & Hawisher, 1990 Kaplan, 1995 digressiveness/linearity relevance directionality textual symmetry grouping; framing rhythm of discourse data integration relationships among items (no parallel) advance organizers framing; metalanguage form vs. content sentence types framing form vs. content tWhile this comparison is an oversimplification of the theories, it does show the amount of overlap that exists in contrastive rhetoric methods.
Figure 1. Comparison of elements identified in contrastive analysis. 246 K. Precht Digression/linearity. A text is determined to be digressive if the LR discusses anything other than the applicant or basic aspects of the educational system that would be critical to understanding the applicant’s performance. An example of a digressive text would be the Eastern European LRJ28-EE (see Appendix A) which moves from discussing the applicant, to describing the political situation in his home country, and then back to discussing the applicant. Any move away from the topic of the applicant or educational system that was longer than one sentence was considered to be a digression. Since the texts were short and narrowly focused, a digression as short as two sentences is noticeable, and thus a reasonable length for determining linearity. Symmetry.
A text is determined to have symmetry if it had short (3-5 line) introductory and concluding paragraphs, and a body paragraph or paragraphs which were longer than the introduction and concluding paragraphs. The American LR M36-US in the appendix is symmetrical, while the German LR G21-G is asymmetrical. Data integration. A text was considered integrated if the accounts of academic success or accomplishments are mentioned to prove a larger point; it was considered non-integrative if the data are listed as a proposition, without interpretation.
The following paragraph, excerpted from the British LR R53-UK in Appendix A, is not integrated: R’s superb First in [exam name] was thoroughly deserved, and a proper reflection of his legal abilities and commitment to his studies. He is an excellent analyst of legal problems, cutting to the heart of the issues presented to him and working thoroughly and logically to the solutions. He writes clearly, concisely and eloquently. His oral skills are considerable and he thinks very quickly and correctly on his feet. [R53-UK] This list of qualities is not used to support a larger topic, and the writer does not refer back to these traits later in proving his point.
The list is presented as evidence without interpretation. Integrated texts, on the other hand, use evidence in support of a larger point: Ms. G is an exceptional student not only by her grades but by the curriculum she put together for herself as well. She combined courses in classics and history with economics and the complete law school program. Therefor [sic], her academic background is much broader and deeper as even a very good law student’s. The project Ms. G is presenting for a Master’s thesis seems well prepared.
The [topic of thesis] bears enough theoretical and practical weight to deserve further research. [G20-G] This text shows the use of evidence to prove a larger point, i. e. that “Ms. G is an exceptional student”. Advance organizers. The text was determined to have advance organizers if A Cross-cultural Comparison of Letters of Recommendation 247 it gave advance notice to the reader of the order of topics to be discussed, or gave other signals to the reader of text organization. As the purpose of advance organizers is to help the reader develop expectations for the format of the text, any mention of the text being a LR was considered to be an advance organizer.
Examples of advance organizers are as follows: i. I am writing to recommend Dr. B for a course leading to a Masters degree in law. [B4-EE] ii. [M] exuded intellectual curiousity and excellence, which he combined with a gracious personal modesty. I will say more about each of these things below. [M35-US] Example (i) is considered an advance organizer in announcing its purpose as a LR, which implies a particular organizational pattern (whether that pattern is indeed what the reader expects, due to cultural differences, is not at issue). Example (ii) classifies the qualities of the applicant, and explicitly states that the LR will be organized around a description of these qualities, so it also qualifies as use of an advance organizer.
Topic, enumerative and bridge sentences. Three types of sentences were examined: Topic sentences were determined to be sentences which began a paragraph and contained the controlling ideas for the paragraph. The German LR G20-G, shown earlier, displays a topic sentence, which gives the controlling idea for the paragraph. An LR was counted as having topic sentences if at least half the paragraphs had topic sentences. Enumerative sentences list the types within a category that were to be examined in a text.
Any LR that listed categories or examples in organizing the content was considered to be enumerative. Example (ii) (mentioned earlier) is an example of an enumerative sentence in that it lists a classification scheme. Bridge sentences link ideas from the end of one paragraph to the beginning of the next. Most often a bridge sentence will be the first sentence, continuing on with the ideas from the previous paragraph, and finally introducing the topic of the new paragraph: Miss G has excellent knowledge in German Law and is especially interested in foreign law and International Law. [G18-G]
This example shows how the first part of the sentence links back to the previous paragraph which discussed the applicant’s success in studying German law, and the second part of the sentence introduces the topic of the new paragraph, international law. An LR was counted as having bridge sentences if at least half the paragraphs had bridge sentences.
Each LR was analyzed to determine the presence of the above features, 248 K. Precht TABLE 2 Digressiveness and Linearity American German British Eastem European N=10 N=9 N=10 N=10 Digressiveness 2 1 3 6 Linearity 8 8 7 4 and the counts below are occurrences of features, out of a total of 9 for Germany, 10 for the rest. The second half of the analysis focused on comparing the structure and content of the LRs. This analysis focused on identifying differences in evidence and development within the frame, body and conclusion of the LRs. The procedures for analysis corresponded to Bhatia’s (Bhatia 1993: 19- 22) Level 3 in genre analysis, where patterns are identified within the requirements of the communicative purpose of the genre. Results and Discussion Digressiveness/linearity. In the LRs, the German, British and American texts were quite similar in digressiveness/linearity (see Table 2).
Similarities in the American, German and British linearity may be due to the length of the texts. Such short, narrowly focused texts afford little opportunity for digression, as compared to the complexity and length of a research article. This finding is interesting in light of Clyne (1991) who suggested that linearity was a sociocultural feature that was especially difficult for German writers to master when writing in English, and found German texts to have more digressions than either British or American texts.
This is an important measure in LRs in that letters that seem digressive could be discounted as irrelevant, and readers from linear backgrounds may not be able to make the connections between the digressions and the larger context. Textual symmetry. The Germans wrote the least symmetrical texts, followed by Eastern European and American texts (see Table 3). In examining the LRs, the asymmetry in texts seems to be the result of other textual considerations. For the Americans, the asymmetry seems to weight the importance of the topic of the paragraphs; personal characteristic TABLE 3 Textual Symmet~ry American German British Eastern European N=10 N=9 N=10 N=10 Symmetry 7 4 9 6 Asymmetry 3 5 1 4
A Cross-cultural Comparison of Letters of Recommendation TABLE 4 Integration vs Nonintegration of Data 249 American German British Eastern European N=10 N=9 N=10 N=10 Integration 10 6 8 8 Nonintegration 0 3 2 2 paragraphs were generally shorter than paragraphs on accomplishments. For the Germans, though, the shorter paragraphs were related to data integration, as paragraph length was in large part due to how well the data were integrated into the text (see discussion on integration that follows).
Eastern European letters show evidence of both weighting and data interpretation, and in addition, often the lengths of paragraphs correspond to the importance of the event described. Data integration. The Americans were more likely to have data integration than any of the other regions (see Table 4). Data integration has been linked to reader vs writer responsibility (Manranen 1993), and as such, American texts would be expected to have more data integration. These results certainly correspond with an American tendency toward data integration. In a LR, the implicit purpose of the text is to evaluate the applicant, and this evaluation is commonly the interpretation of “data”–the accomplishments of the applicant.
Since interpretation of the data is one definition of data integration, it would stand to reason that there would be more data integration cross-regionally than perhaps might be used in other texts. It’s worth noting that the two British LRs that did not exhibit data integration both used a format which begins with a list of the applicant’s accomplishments, followed by the writer’s opinion of the applicant. Advance organizers and sentence types. Advance organizers and the sentence types–topic, enumerative and bridge–work as a metalinguistic cue to the reader in organizing the information in the text.
These differences show that, while each region has signals for the reader in the text, these metalinguistic cues vary regionally. The data in Table 5 certainly fall in line with characterizations of American writing as the five-paragraph theme, using a great many topic sentences and TABLE 5 Advance Organizers and Sentence Types (number of occurrences ) American German British Eastern European N=10 N=9 N=10 N=10 Advance organizers 9/10 3/9 5/10 5/10 Topic sentences 7/10 4/9 5/10 3/10 Enumerative sentences 2/10 3/9 0/10 1/10 Bridge sentences 0/10 5/9 3/10 3/10 250 ~ Precht advance organizers (see Popken 1987; Clyne 1991).
The task of the LR required sentence-level organizational features for the American writers: most American writers categorized applicant characteristics and developed the categories in separate paragraphs, which were each introduced by topic sentences. Table 5 indicates that other regions consistently use sentencelevel text organizers as well. The relatively high use of bridge sentences by Germans, confirms earlier studies which found Germans more likely to use bridge sentences than topic sentences (Clyne 1991).
The somewhat low sentence-level organizers on the part of the British could in part be explained by the tendency for some British writers to begin with a list of accomplishments rather than with an introduction-thesis statement, as was noted in data-integration, earlier. Structural Patterns All LRs were written in an introduction/body/conclusion format. The introduction served as a frame for the LR, the body contained the main evaluation of the applicant, and the conclusion contained predictions of success for the applicant. Beyond this broad structure, though, there was a great deal of regional variation.
The greatest differences cross-culturally were in the organizational patterns and in the methods of support used by the writers. Frame. The frame was defined to be the introductory remarks of the letter that prefaced the body of the letter, and the frame was commonly set apart in its own paragraph. In the frame, the writers explain the reason they are writing: to recommend the applicant. Even the Eastern European LRs, which are greatly context-reduced, had clues by which they announced the intention of the writer. Stating the purpose is a type of advance organizer, as in Table 5. There are three main elements to the frame: the explicitness with which the purpose for writing is stated, whether the context in which the recommender knows the applicant is stated, and the degree of distance or warmth the recommender displays.
Frequencies of these elements of the frame per region are shown in Table 6. TABLE 6 Frequency of Elements in The Frame of Letters of Recommendationt Purpose stated Context of knowing Personal comments in frame applicant stated in frame included in frame American 7 9 1 British 5 7 4 German 3. 3 7. 7 8. 8 Eastern European 3 10 5 tNumber of occurrences out of a total of 10 letters, with German letters normed to 10. A Cross-cultural Comparison of Letters of Recommendation 251 American. The American writers immediately justify their reasons for writing.
In the frame, the writers often explicitly state that they are writing a letter of recommendation; seven of the 10 LRs explicitly stated that the writer was recommending a student for admission, with an opening statement such as: i. I have been asked to write in support of the application of… [B5-US] ii. t write to give [M] the highest possible recommendation… [M37-US] Sometimes the expression of purpose becomes something of an overt speech act in some of the American LRs, as in (ii). Nine of the 10 American writers included the context in which they know the applicant in the frame.
The personal content of the relationship is often stressed either through the content itself or through possessive pronouns which show the writer’s relationship to the applicant. I know Mr. N very well after working with him for more than a year. [N39- us] Mr. Jwas a student of mine last summer… ~25-US] (italics added) British. The British letters were split between those which began with a list of facts with little data integration, and those that began with introductory remarks and a “thesis statement” of sorts for the letter.
Those that began with a list of facts make no mention of the context in which the writer knew the applicant, or the purpose for which the letter was written; these LRs begin abruptly with a litany of the applicants’ achievements, more like what one expects to see in the body/evaluation section: R’s superb First in [name of exam] was thoroughly deserved, and a proper reflection of his legal abilities and commitment to his studies. He is an excellent analyst of legal problems… [R53-UK] Furthermore, these texts have a great deal of evaluation in the opening paragraph: His record of seven first class marks out of nine in the two years is without question extremely good. [R51-UK] [emphasis his]
The rest of the British letters begin exactly as the American LRs did, mentioning the purpose for writing and often giving an overt speech act recommendation: I write in support of the application made by Mr. S to undertake study in the LLM program. [S54-UK] German. German frames are unique in three respects. First, they lack a 252 I~ Precht direct reference to the purpose of the letter. The German LRs state the purpose for writing only one-third of the time (three of nine LRs), and none of the LRs included an overt speech act recommendation in the frame.
Second, the frames contain direct references to context in which the writer knows the applicant, and describe how the appficant came to distinguish herself or himself, usually through writing exceptional papers or exams. Third, the frames include an evaluation of the applicant. Eight of the 9 frames of German LRs contained an evaluation of some aspect of the applicant, quite often as a part of the context in which the writer knew the applicant: C caught my attention by submitting excellent papers in a beginners course on private law. [C7-G] I know Miss G well. She was one of the best students in my lessons of civil law and civil procedural law. [G18-G] Eastern Europe.
The context in which the writer knows the applicant was given considerably more emphasis in the frame, and is almost invariably the first statement of the letter. The extent of the relationship with the applicant is emphasized; all 10 LRs related the context in which the writer knows the applicant. In the Eastern European personal narrative style, the frame serves to give background information before the narrative proper. While the context descriptors are quite similar to the German knowledge-of-applicant statements, in the Eastern European LRs they serve as springboards to begin the story of how the writer has come to know the applicant.
I have known Mr. J since he became a student at post-graduate studies at the (university name), which I teach courses in …. 26-EE] I have known Mr. J for a long time, first as a graduate student …. in more recent times as a collaborator. [J28-EE] Eastern European writers are likely to evaluate the applicant in the frame; half the writers included evaluations in their frames. These evaluations follow directly on the heels of the contextualizing statements, and serve as an advance organizer for the narrative: Mention has also to be made of his outstanding working capabilities, and
reliability in meeting his commitments. [J27-EE] P is one of the most intelligent, insightful and perceptive students I have known at (university name). [P47-EE] Eastern European texts rarely state the purpose for writing. Of the 10 LRs, only three have an explicit mention of the purpose in the frame. Body/evaluation. The body of all the letters was mainly evaluative, though evaluation was expressed quite differently from one region to another. There A Cross-cultural Comparison of Letters of Recommendation TABLE 7 Topicality vs Chronology and Listing of Facts vs Storytellin