Writer Perception, Writer Projection: The Influence of Personality, Ideology, and Gender on Letters of Recommendation
by Henrik Eger
Ph.D. Dissertation, University of Illinois at Chicago, 1991, 328 pages.
by Henrik Eger
Ph.D. Dissertation, University of Illinois at Chicago, 1991, 328 pages.
Abstract: The influence of personality, ideology, and gender on the process of composing letters of recommendation (LORs) in the American business culture is investigated.
Thesis: Writers are so restrained by their gender, ideology (deeply held, mainly unexamined values, attitudes, and world views), and personality (Meyers-Briggs Type Indicator), that they will compose LORs based on criteria which confirm their own subjective sense of self, rather than write texts which approximate a candidate's reality.
Thesis: Writers are so restrained by their gender, ideology (deeply held, mainly unexamined values, attitudes, and world views), and personality (Meyers-Briggs Type Indicator), that they will compose LORs based on criteria which confirm their own subjective sense of self, rather than write texts which approximate a candidate's reality.
Drawing of wings by Leonardo da Vinci, 1452-1519. Dissertation dedication by Henrik Eger, Ph. D., June 28, 1990.
Results: Each LOR carried, Trojan horse-like, some of the writer's concealed subtexts. If a candidate was seen as similar or identical, most writer's ranked that person higher, used fewer negative words, and committed fewer genre violations than for a candidate whose profile did not match theirs. Perception of difference in ideology caused strong, adverse projections and almost total rejection. The least educated writers (measured by percentage of grammatical/stylistic problems) violated genre conventions 84% more often than the most educated writers.
WRITER PERCEPTION, WRITER PROJECTION: THE INFLUENCE OF PERSONALITY, IDEOLOGY, AND GENDER ON LETTERS OF RECOMMENDATION
SUMMARY
The influence of personality, ideology, and gender on the process of composing letters of recommendation (LORs) in the American business culture is investigated. Thesis: Writers are so restrained by their gender, ideology (deeply held, mainly unexamined values, attitudes, and world views), and personality (Myers-Briggs Type Indicator), that they will compose LORs based on criteria which confirm their own subjective sense of self, rather than write texts which approximate a candidate’s reality.
The study generated 144 LORs written by 12 male and 12 female professionals who wrote LORs on five hypothetical candidates whose educational background and job performance were identical. Their personality and ideology matched that of the writers (2 LORs) and was the opposite (2 LORs). Only neutral data were supplied for the fifth candidate. The writers wrote the sixth LOR about themselves. Independent variables were personality, ideology, and gender. Dependent variables were linguistic features: words, negative words, passive voice, grammatical/stylistic problems, readability, descriptive level, genre violations (negative, misleading, or illegal statements), and writers’ analytical comments.
The results confirmed the hypothesis that professionals perceive and judge others based on their own personality, ideology, and gender. Each LOR carried, Trojan horse-like, some of the writer’s concealed subtexts. If a candidate was seen as similar or identical, most writers ranked that person higher, used fewer negative words, and committed fewer genre violations than for a candidate whose profile did not match theirs. Knowledge of candidates’ gender led to preferential treatment of male candidates by male professionals and preferential treatment of female candidates by female professionals (“advocacy effect”). Identification with candidates’ personality led to very positive comments and the highest ranking. Perception of difference in ideology caused strong, adverse projections and almost total rejection. The least educated writers (measured by percentage of grammatical/stylistic problems) violated genre conventions 84% more often than the most educated writers. A study of the writers’ analytical comments about their own LORs indicated that most writers were surprised about their deeply-seated bias.
The study concludes that professionals need to practice an active acceptance of difference, based on the belief that difference among people can lead to human and professional growth through its very challenge.
1. INTRODUCTION
We measure the excellency of other men [women and children] by some excellency we conceive to be in ourselves.
—John SELDEN, “Measure of Things,” Table Talk (1689)
In each action [each letter of recommendation] we must look beyond the action [beyond the letter of recommendation] at our past, present, and future state, and at others whom it affects, and see the relations of all those things. And then we shall be very cautious.
—Blaise PASCAL, Pensées, (1670), 505, tr. W. F. Trotter
Read not to contradict and confute, nor to believe and take for granted, nor to find talk and discourse, but to weigh and consider.
—Francis BACON, Of Studies (1625)
1.1 Background
Some years ago, I studied toward an M.A. in Sociology of Literature at the University of Essex at Colchester in England. The main emphasis was on ideology and literary theories. As a result of my studies, I wrote a thesis entitled “Distortions: A Socio-Literary Study of some Influences on the Production and Perpetuation of the Image of the German in English Literature” (1976). In it, I presented evidence to suggest the possibility that perceptions of others are directly related to external events, ruling ideologies, and the general Zeitgeist.
For example, during the reign of Queen Victoria, whose husband was Prince Albert of Saxony, British authors tended to describe German characters as human beings who were not only outstandingly bright and hard working, but also role-models for British readers. However, by the time World War I broke out, most British authors had changed their use of language dramatically and described people who happened to live in Germany as sub-human, labeled them “Huns,” and reduced them to “swine” who needed nothing but a British bayonet in their body. J. A. Hammerton, the English editor of The Masterpiece Library of Short Stories: The Thousand Best Complete Tales of All Times and All Countries (London, n.d. [circa 1923]), said in his introduction that nothing written in German was worth being included in the 22 volume international collection.
Katherine Mansfield, in a letter to J. M. Murray, was the only author found who actually retracted her creations and who openly admitted that the German characters in her short stories (In A German Pension) were “lies.” Authors like D. H. Lawrence and A. Huxley are among the few British writers during the same time period who presented complex German characters with very human characteristics (see Eger, “The Image of the German in Selected English Short Stories, 1809-1974,” 1974).
However, a review of short stories written after World War I strongly suggested that a perception of ideological difference between “the British” and “the Germans” by many British writers led to an increase in the creation of one-sided characters and dichotomies of good and evil—at the expense of more complex characters. Similar studies could be done with the image of the English (“Tommies”) in German literature, or the image of North Americans (“gringos”) in Latin American literature, etc. Most likely, the general perceptions of the other at the time influences the majority of writers who, in turn, will create texts reflective of ruling ideologies.
If the main emphases of my studies in Europe were on socio-literary aspects of text production, the main interests of my recent study are psychological aspects of writing. I have been particularly influenced by the Carl Jung-inspired works of Isabel Briggs Myers and Katherine C. Briggs. I also owe much to the publications of George H. Jensen and John K. DiTiberio on personality and composing processes. All the above authors encourage writers, teachers, and students of writing to recognize and accept difference. As a result of such insights, teachers of composition can now work with much more understanding of their own and their students’ composing processes.
Several other aspects of life in the United States led me to develop the present study, especially the strong emphasis on hard research data and the general interest in America on the practical. In short, my studies over the past 15 years in different cultures encouraged me to bring together some of the theory-driven work which I had experienced in Europe with some of the empirical, research-based work that I had experienced in the United States.
Last, but not least, this study owes its existence to an interest in a phenomenon that seems to bridge cultures and times: many intelligent and well-educated people who appear rational and considerate of others frequently present shifting and often conflicting interpretations of people and events. They tend to view one and the same person, one and the same action, in radically different ways. What was “outstanding” and “very special” can quickly become “dreadful” and “threatening.”
As a result, I began to investigate the possibility that “reality” is not a given but a construction, made up of a series of interacting and constantly changing perceptions and projections. I began to wonder why such a construction through perception and projection can take on a whole new “truth dimension” which did not exist before and why a new “reality” can be created with each speech act, each letter of recommendation (hereafter referred to as LOR), adding to the multitude of already existing “realities” of just one person. In short, I replaced the concept of “reality” with the more flexible concept of “reality construction” which allows for a multitude of layers of perceptions and projections by many different individuals. It is against this background that this study evolved.
1.2 Statement of the Problem
Comparatively little is known in rhetoric and composition research about the influence of personality type on the process of writing; even less is known about knowledge—real or perceived—that writers have of candidates who have requested an LOR and the influence of that knowledge on the production of text; and apparently nothing concrete is known about the influence of personality types, values, ideologies, and performance data on the actual production of LORs, especially the consequences on the process of writing LORs when writers perceive candidates as having quite a different set of values and ideologies from that of the writers. Yet thousands of professionals and expert writers (hereafter called “writers”)—people who write as part of their profession, for example, executives, managers, business people, heads of departments, or instructors—sit down many times during the year in their business, government, or school offices to compose persuasive LORs to other professionals in the field. They write important documents that frequently go beyond the level of skills and performance description into areas which can strongly influence the readers in their decision making.
While it may be obvious to most writers that the process of writing is usually highly goal oriented (Hayes and Flower, 1986), not all writers may be aware that they could be operating under the illusion that their LORs are rational and objective tools which measure and transmit important, verifiable data on job performance and skills. Instead, writers may be guided by data selection criteria that have to do as much with their own personality types, ideologies, and gender subtexts as with those of the employee or recommendee (hereafter called “candidate”) and the intended target audience. If this assumption is true, then the process of writing LORs inevitably results in highly subjective pieces of writing, in documents that can reveal as much about the writer as they can conceal about the candidate. However, these very problems—deeply embedded in a text—can also lead to new insights.
This study investigates aspects of some of the problems caused by writer perception and projection, especially the influence of personality types and ideologies on the production of LORs. I will demonstrate at least part of the concealed “problematic” (an ambiguous or problematic relationship) of this type of discourse production, namely “interpersonal affect and rating errors” (Barry and Tsui, 1986), that is, the tendency of writers and “raters [to] prefer candidates who are similar to themselves” (Glick, 1981). This selective treatment of candidates by fellow professionals—based on a sense of writer-identity with a candidate’s personality, ideology, and gender—seems to guide part of the composition process and employment selection. Yet, the influences of writer identification—or non-identification—with a candidate and their consequences for the production of texts appear almost unexplored.
1.3 Significance of the Problem
The act of persuasion has been of interest to speakers and writers for thousands of years. During the last twenty years a renewed interest in composition and rhetoric, especially in the United States, has helped to create a growing body of knowledge in persuasive writing and discourse production as witnessed by an ever-increasing number of dissertations, books, and articles, including a large number of studies that deal with LORs (see Bibliography). Similarly, the following areas are contributing to an understanding of some of the phenomena of writer perception and writer projection: contemporary literary theories, especially deconstructionism, with their attempt to come to terms with subtexts or the “unwritten text”; Marxist-based studies with their emphasis on the importance of ideologies on the production of text; and educational, psychological, and business-related studies, especially those that deal with values and personality types.
I determined that a truly interdisciplinary approach was necessary to develop this research project, in order to come to a better understanding of a significant problem, namely the hidden problematic of LORs in the American business culture. I hope that such a cross-disciplinary approach will help make visible some of the perceived discrepancies between identical sets of performance data on candidates and widely-diverging texts which seem to say at least as much about the writers as they say about the candidates.
1.4 Purpose of the Study
One of the main purposes of this study is to investigate unknown aspects of text production, especially the production of LORs. It does not set out to dethrone LORs or even to doubt the usefulness of LORs. Rather, based on the data generated through several writing experiments, this study questions uncritical or one-sided acceptance or rejection of LORs. It builds toward an understanding of text production through a model which allows for multiple reality constructions—realities that include or allow for perceptions and projections of each writer, candidate, and reader. It is suggested that these multiple layers of reality coexist—shifting and changing from encounter to encounter, from moment to moment. With this model of “reality construction,” one can possibly see better at least some layers of previously hidden subtexts. Yet, so far, quite a few of these aspects of text production seem to have been neglected or ignored in the literature on LORs. Ultimately, a deeper knowledge and understanding of writer perception and writer projection could possibly lead to changes in the perception of job applicants and a better understanding of LORs and the writers of such documents.
Therefore, this study sets out to combine important work done in various fields in order to contribute to the ongoing research in text production. If the main hypotheses are proved, the results might help writers and readers of LORs to better understand aspects of the processes that can influence their writing. Future studies might even lead toward the development and assessment of a “grammar of acceptance and rejection,” provided the rhetorical context is known, based on the individual perceptions of each other’s personalities, ideologies, and gender expectations.
Above all, it is hoped that this study will contribute to the ongoing research in discourse production. It is guided by the hope that better understanding of composition processes will eventually contribute to a greater acceptance of difference, rather than rejection. I also hope that an active awareness of data selection processes will help writers overcome deeply seated attitudes and behavioral patterns and eventually lead to a greater acceptance of people who are different in their personality, ideology, or gender.
*For more excerpts of this dissertation, or to read the entire document, contact the editor.
SUMMARY
The influence of personality, ideology, and gender on the process of composing letters of recommendation (LORs) in the American business culture is investigated. Thesis: Writers are so restrained by their gender, ideology (deeply held, mainly unexamined values, attitudes, and world views), and personality (Myers-Briggs Type Indicator), that they will compose LORs based on criteria which confirm their own subjective sense of self, rather than write texts which approximate a candidate’s reality.
The study generated 144 LORs written by 12 male and 12 female professionals who wrote LORs on five hypothetical candidates whose educational background and job performance were identical. Their personality and ideology matched that of the writers (2 LORs) and was the opposite (2 LORs). Only neutral data were supplied for the fifth candidate. The writers wrote the sixth LOR about themselves. Independent variables were personality, ideology, and gender. Dependent variables were linguistic features: words, negative words, passive voice, grammatical/stylistic problems, readability, descriptive level, genre violations (negative, misleading, or illegal statements), and writers’ analytical comments.
The results confirmed the hypothesis that professionals perceive and judge others based on their own personality, ideology, and gender. Each LOR carried, Trojan horse-like, some of the writer’s concealed subtexts. If a candidate was seen as similar or identical, most writers ranked that person higher, used fewer negative words, and committed fewer genre violations than for a candidate whose profile did not match theirs. Knowledge of candidates’ gender led to preferential treatment of male candidates by male professionals and preferential treatment of female candidates by female professionals (“advocacy effect”). Identification with candidates’ personality led to very positive comments and the highest ranking. Perception of difference in ideology caused strong, adverse projections and almost total rejection. The least educated writers (measured by percentage of grammatical/stylistic problems) violated genre conventions 84% more often than the most educated writers. A study of the writers’ analytical comments about their own LORs indicated that most writers were surprised about their deeply-seated bias.
The study concludes that professionals need to practice an active acceptance of difference, based on the belief that difference among people can lead to human and professional growth through its very challenge.
1. INTRODUCTION
We measure the excellency of other men [women and children] by some excellency we conceive to be in ourselves.
—John SELDEN, “Measure of Things,” Table Talk (1689)
In each action [each letter of recommendation] we must look beyond the action [beyond the letter of recommendation] at our past, present, and future state, and at others whom it affects, and see the relations of all those things. And then we shall be very cautious.
—Blaise PASCAL, Pensées, (1670), 505, tr. W. F. Trotter
Read not to contradict and confute, nor to believe and take for granted, nor to find talk and discourse, but to weigh and consider.
—Francis BACON, Of Studies (1625)
1.1 Background
Some years ago, I studied toward an M.A. in Sociology of Literature at the University of Essex at Colchester in England. The main emphasis was on ideology and literary theories. As a result of my studies, I wrote a thesis entitled “Distortions: A Socio-Literary Study of some Influences on the Production and Perpetuation of the Image of the German in English Literature” (1976). In it, I presented evidence to suggest the possibility that perceptions of others are directly related to external events, ruling ideologies, and the general Zeitgeist.
For example, during the reign of Queen Victoria, whose husband was Prince Albert of Saxony, British authors tended to describe German characters as human beings who were not only outstandingly bright and hard working, but also role-models for British readers. However, by the time World War I broke out, most British authors had changed their use of language dramatically and described people who happened to live in Germany as sub-human, labeled them “Huns,” and reduced them to “swine” who needed nothing but a British bayonet in their body. J. A. Hammerton, the English editor of The Masterpiece Library of Short Stories: The Thousand Best Complete Tales of All Times and All Countries (London, n.d. [circa 1923]), said in his introduction that nothing written in German was worth being included in the 22 volume international collection.
Katherine Mansfield, in a letter to J. M. Murray, was the only author found who actually retracted her creations and who openly admitted that the German characters in her short stories (In A German Pension) were “lies.” Authors like D. H. Lawrence and A. Huxley are among the few British writers during the same time period who presented complex German characters with very human characteristics (see Eger, “The Image of the German in Selected English Short Stories, 1809-1974,” 1974).
However, a review of short stories written after World War I strongly suggested that a perception of ideological difference between “the British” and “the Germans” by many British writers led to an increase in the creation of one-sided characters and dichotomies of good and evil—at the expense of more complex characters. Similar studies could be done with the image of the English (“Tommies”) in German literature, or the image of North Americans (“gringos”) in Latin American literature, etc. Most likely, the general perceptions of the other at the time influences the majority of writers who, in turn, will create texts reflective of ruling ideologies.
If the main emphases of my studies in Europe were on socio-literary aspects of text production, the main interests of my recent study are psychological aspects of writing. I have been particularly influenced by the Carl Jung-inspired works of Isabel Briggs Myers and Katherine C. Briggs. I also owe much to the publications of George H. Jensen and John K. DiTiberio on personality and composing processes. All the above authors encourage writers, teachers, and students of writing to recognize and accept difference. As a result of such insights, teachers of composition can now work with much more understanding of their own and their students’ composing processes.
Several other aspects of life in the United States led me to develop the present study, especially the strong emphasis on hard research data and the general interest in America on the practical. In short, my studies over the past 15 years in different cultures encouraged me to bring together some of the theory-driven work which I had experienced in Europe with some of the empirical, research-based work that I had experienced in the United States.
Last, but not least, this study owes its existence to an interest in a phenomenon that seems to bridge cultures and times: many intelligent and well-educated people who appear rational and considerate of others frequently present shifting and often conflicting interpretations of people and events. They tend to view one and the same person, one and the same action, in radically different ways. What was “outstanding” and “very special” can quickly become “dreadful” and “threatening.”
As a result, I began to investigate the possibility that “reality” is not a given but a construction, made up of a series of interacting and constantly changing perceptions and projections. I began to wonder why such a construction through perception and projection can take on a whole new “truth dimension” which did not exist before and why a new “reality” can be created with each speech act, each letter of recommendation (hereafter referred to as LOR), adding to the multitude of already existing “realities” of just one person. In short, I replaced the concept of “reality” with the more flexible concept of “reality construction” which allows for a multitude of layers of perceptions and projections by many different individuals. It is against this background that this study evolved.
1.2 Statement of the Problem
Comparatively little is known in rhetoric and composition research about the influence of personality type on the process of writing; even less is known about knowledge—real or perceived—that writers have of candidates who have requested an LOR and the influence of that knowledge on the production of text; and apparently nothing concrete is known about the influence of personality types, values, ideologies, and performance data on the actual production of LORs, especially the consequences on the process of writing LORs when writers perceive candidates as having quite a different set of values and ideologies from that of the writers. Yet thousands of professionals and expert writers (hereafter called “writers”)—people who write as part of their profession, for example, executives, managers, business people, heads of departments, or instructors—sit down many times during the year in their business, government, or school offices to compose persuasive LORs to other professionals in the field. They write important documents that frequently go beyond the level of skills and performance description into areas which can strongly influence the readers in their decision making.
While it may be obvious to most writers that the process of writing is usually highly goal oriented (Hayes and Flower, 1986), not all writers may be aware that they could be operating under the illusion that their LORs are rational and objective tools which measure and transmit important, verifiable data on job performance and skills. Instead, writers may be guided by data selection criteria that have to do as much with their own personality types, ideologies, and gender subtexts as with those of the employee or recommendee (hereafter called “candidate”) and the intended target audience. If this assumption is true, then the process of writing LORs inevitably results in highly subjective pieces of writing, in documents that can reveal as much about the writer as they can conceal about the candidate. However, these very problems—deeply embedded in a text—can also lead to new insights.
This study investigates aspects of some of the problems caused by writer perception and projection, especially the influence of personality types and ideologies on the production of LORs. I will demonstrate at least part of the concealed “problematic” (an ambiguous or problematic relationship) of this type of discourse production, namely “interpersonal affect and rating errors” (Barry and Tsui, 1986), that is, the tendency of writers and “raters [to] prefer candidates who are similar to themselves” (Glick, 1981). This selective treatment of candidates by fellow professionals—based on a sense of writer-identity with a candidate’s personality, ideology, and gender—seems to guide part of the composition process and employment selection. Yet, the influences of writer identification—or non-identification—with a candidate and their consequences for the production of texts appear almost unexplored.
1.3 Significance of the Problem
The act of persuasion has been of interest to speakers and writers for thousands of years. During the last twenty years a renewed interest in composition and rhetoric, especially in the United States, has helped to create a growing body of knowledge in persuasive writing and discourse production as witnessed by an ever-increasing number of dissertations, books, and articles, including a large number of studies that deal with LORs (see Bibliography). Similarly, the following areas are contributing to an understanding of some of the phenomena of writer perception and writer projection: contemporary literary theories, especially deconstructionism, with their attempt to come to terms with subtexts or the “unwritten text”; Marxist-based studies with their emphasis on the importance of ideologies on the production of text; and educational, psychological, and business-related studies, especially those that deal with values and personality types.
I determined that a truly interdisciplinary approach was necessary to develop this research project, in order to come to a better understanding of a significant problem, namely the hidden problematic of LORs in the American business culture. I hope that such a cross-disciplinary approach will help make visible some of the perceived discrepancies between identical sets of performance data on candidates and widely-diverging texts which seem to say at least as much about the writers as they say about the candidates.
1.4 Purpose of the Study
One of the main purposes of this study is to investigate unknown aspects of text production, especially the production of LORs. It does not set out to dethrone LORs or even to doubt the usefulness of LORs. Rather, based on the data generated through several writing experiments, this study questions uncritical or one-sided acceptance or rejection of LORs. It builds toward an understanding of text production through a model which allows for multiple reality constructions—realities that include or allow for perceptions and projections of each writer, candidate, and reader. It is suggested that these multiple layers of reality coexist—shifting and changing from encounter to encounter, from moment to moment. With this model of “reality construction,” one can possibly see better at least some layers of previously hidden subtexts. Yet, so far, quite a few of these aspects of text production seem to have been neglected or ignored in the literature on LORs. Ultimately, a deeper knowledge and understanding of writer perception and writer projection could possibly lead to changes in the perception of job applicants and a better understanding of LORs and the writers of such documents.
Therefore, this study sets out to combine important work done in various fields in order to contribute to the ongoing research in text production. If the main hypotheses are proved, the results might help writers and readers of LORs to better understand aspects of the processes that can influence their writing. Future studies might even lead toward the development and assessment of a “grammar of acceptance and rejection,” provided the rhetorical context is known, based on the individual perceptions of each other’s personalities, ideologies, and gender expectations.
Above all, it is hoped that this study will contribute to the ongoing research in discourse production. It is guided by the hope that better understanding of composition processes will eventually contribute to a greater acceptance of difference, rather than rejection. I also hope that an active awareness of data selection processes will help writers overcome deeply seated attitudes and behavioral patterns and eventually lead to a greater acceptance of people who are different in their personality, ideology, or gender.
*For more excerpts of this dissertation, or to read the entire document, contact the editor.
