Dr. Jeff Sheldon on Channeling Your Inner Hunter S. Thompson: Ethnography, Gonzo Evaluation, and the Nature of Reality

 Lately I’ve been reading the 2016 edition of “Paper Lion,” George Plimpton’s epic account of his foray into the world of professional football during the Detroit Lion’s 1965 pre-season training camp. Plimpton - American journalist, writer, literary editor - was famous not only for his patrician demeanor (Exeter, Harvard), friendship with Robert F. Kennedy, and editorship of the Paris Review, but for his participatory, immersive journalism, taking part in sporting events [his other professional sports writing included baseball (Out of My League, 1961), golf (The Bogey Man, 1967), boxing (Shadow Box, 1977), and ice hockey (Open Net, 1985)]; acting in a Western; performing a comedy act at Caesars Palace in Las Vegas; and playing with the New York Philharmonic Orchestra; it has been said of Plimpton that he insinuated himself with ease into the myriad roles he undertook. This is not to say he was good at any of them, but he took them on with a certain panache. The journalistic aspect of these escapades was Plimpton’s reportage on each experience from the point of view of the amateur engaged in a professional pursuit, which, ultimately, he turned into his highly regarded books and articles. According to acclaimed American author Nicholas Dawidoff, who wrote the forward for this edition of Paper Lion:

“By inventing the participatory, immersive journalism genre Plimpton gave otherwise obstructed observers the means for seeing distant subjects up close.”

As I’ve read with fascination and great enthusiasm about Plimpton’s transformation into the Lion’s “last-string” quarterback through his immersion and participation in every aspect of Lion’s training camp, it got me thinking about author/journalist Hunter S. Thompson (e.g., 1971’s Fear and Loathing in Las Vegas, 1973’s Fear and Loathing on the Campaign Trail’72, and numerous Rolling Stone and ESPN articles). Just as I’ve been enjoying Plimpton so too have I enjoyed reading Thompson, but for a different reason: his unique brand of journalism, “Gonzo Journalism,” which he coined to describe his style of research and reportage. Notably, Plimpton and Thompson used a similar form of participation and immersion in their journalistic (mis)adventures, and were, somewhat surprisingly given noticeable differences in temperament (and Thompson’s prodigious, illicit drug intake), the very best of friends. Quoting from an undated piece Thompson wrote for ESPN’s Page 2:

“George Plimpton was about as good of a friend as a man can have in this world. He lived his life like a work of fine art. George Plimpton was a winner. He was comfortable with everything, from reading Plato in the original Greek, to sparring with Muhammad Ali and courting Jackie Kennedy. He was an athlete and a scholar. He played touch football with Bobby Kennedy on the lawn of Hickory Hill and built some of the most dangerous and colossal firebombs ever seen in the American Century. He was absolutely fearless. I loved George, and he has been a gigantic influence in my life.”

The above quotation illustrates the nature of the Plimpton – Thompson relationship, but more important the profound influence Plimpton had on Thompson’s more intensive, experiential approach to journalism; if you understand and appreciate Plimpton then you’ll likely understand and appreciate Thompson. That said, the main thrust of this piece is about how one might apply Gonzo Journalism to social science research and evaluation. For the uninitiated, Gonzo Journalism was a style of journalism that claimed little if any objectivity because Thompson became part of the story via a first-person narrative of events in which he participated, extensively using his own experiences and emotions to color the story he was attempting to follow. A prime example of Gonzo Journalism, and the first Thompson book I ever read is his epic 1966 work, Hell's Angels: The Strange and Terrible Saga of the Outlaw Motorcycle Gang which detailed Thompson's experience living with that infamous, mostly notorious motorcycle club founded in San Bernardino, California in 1948. What makes this book a prime example of Gonzo Journalism is that Thompson spent over a year embedded with the Oakland Hell’s Angels chapter learning their unique subculture through full immersion and participation in their lifestyle. According to Leo Litwak in his contemporaneous 1967 New York Times book review, “Hell’s Angels,” Thompson related how he drank at their bars, exchanged home visits, recorded their brutalities, viewed their sexual caprices, converted to their motorcycle mystique, and was so intrigued by their lifestyle he was no longer sure whether he was doing research on the Hell's Angels or being slowly absorbed by them. The difference between Plimpton and Thompson, and what I think makes Thompson’s journalistic style Gonzo, is the extremes to which he went to become a bona fide Hell’s Angel, in part due to the actual length of time he spent in their company, and in part to allowing his emotions and past experiences to influence his reporting. While it’s true that Plimpton immersed himself and participated as an amateur in the professional milieu, his forays were relatively short-term and I get the impression he held something back in that he didn’t allow much, if any, of his emotions and past experiences to influence what and how he wrote. As a result, Plimpton’s reportage seems to me to be more dispassionate and distant than Thompson’s which to me is immediate, proximate, and far more intense. Each knew their respective subject, but the question is which one really understood his respective subject? Defined, knowledge is a familiarity or awareness of someone or something (e.g., facts, information, descriptions, skills, etc.) acquired through experience or an education that includes perceiving, discovering, or learning. Understanding is a step beyond knowledge in that it has been described as a relationship between the knower (in this case Plimpton and Thompson) and the object of understanding. When someone comprehends information learned (i.e., knowledge) then it is said to be understood. Knowledge and understanding are each on a continuum. One can know nothing or a lot about a thing. Likewise one can understand nothing or a lot about a thing. Thompson and Plimpton both knew a lot about their respective subjects from participation and immersion, but Thompson, on the understanding continuum and by virtue of the fact that his experience was longer in duration, more intensive, and more affective better understood the subjects about which he wrote compared to those of Plimpton.          

For the remainder of this piece I connect Thompson’s Gonzo Journalism to the social sciences, arguing for a new type of evaluation that blends objective research methods with subjective, affective experiences to both know and understand one’s research subject. Although Thompson didn’t call his journalistic style ethnography, or himself an ethnographer, the similarities to that branch of the social sciences as both research method and end result are striking. By definition, ethnography is a systematic study of people and cultures designed to explore cultural phenomena in which the researcher, or in this case, journalist, observes and writes about a society from the point of view of the subject of the study through complete immersion and participation within that society. Society, writ large, can either mean an aggregate of people living together in an ordered community or an organization formed for a specific purpose/activity. In Thompson’s case, the Hell’s Angels can certainly be construed as a society irrespective of the nature of its purpose (i.e., motorcycles) or the nefariousness of its activities (i.e., arguably, mischief and mayhem). A brief primer on the mechanics of ethnography is warranted. As a method of data collection ethnography entails examining the behaviour of the participants in a certain specific social situation while trying to understand how they interpret their own behaviour. Dewan (2018) further elaborates that this behaviour may be shaped by the constraints participants feel because of the situations they are in or by the society to which they belong. Ethnography, as a presentation of empirical data on human societies and cultures was pioneered in the biological, social, and cultural branches of anthropology, but also became popular in the social sciences more generally - sociology, communication studies, psychology, economics etc. - wherever people study ethnic groups, formations, compositions, resettlements, social welfare characteristics, materiality, spirituality, and a people's ethnogenesis (i.e., the formation and development of an ethnic group). The typical ethnography is a holistic study including a brief history, an analysis of the terrain, the climate, and habitat. In all cases, it should be reflexive, make a substantial contribution toward the understanding of the social life of humans, have an aesthetic impact on the reader, and express a credible reality as did Thompson. An ethnographer records all observed behavior and describes all symbol-meaning relations, using concepts that avoid causal explanations. Dewan continues that traditionally, ethnography was focused on the far, 'exotic' east, but researchers are now undertaking ethnography in their own social environment (e.g., the Hell’s Angels Oakland Chapter’s clubhouse). Thompson’s resulting books and articles, clearly grounded in ethnography, reflected the knowledge and system of meanings in the lives of the Hell’s Angels as a cultural group and by sharing their lived experience he was better able to interpret their interpretations of their own behaviour.  What set Thompson apart is that he connected so seamlessly with them as different as he might have appeared when he first gained an entrée into their world. As Dewan posits, there are many facades of ourselves that connect us to people and other facades that highlight our differences; in Thompson’s case it was surely the former.     

To my way of thinking, and to that of Thompson I would imagine, if you really want to capture the true essence of your subject (i.e., knowing and understanding it), including all its subtleties and nuances, then Gonzo Journalism is the way to go. In reading Plimpton and reflecting on Thompson it got me thinking about my work as a researcher who conducts evaluations, and how in trying to be completely objective and free of bias I have always distanced myself to the greatest extent possible from the evaluation subjects (i.e., those whose behaviors, knowledge, perceptions, feelings, etc. I’m interested in knowing) and the evaluation context (i.e., where the evaluation takes place in an organization, a community, a village, etc.). However, in musing about objectivity, subjectivity, and the nature of reality, if, as Thompson both posited and exemplified you want to know about another person’s lived experience (including people in the aggregate constituting both types of society) so you can describe it accurately then you have to inhabit that other person’s life, in essence becoming that person, feeling what they feel, learning and knowing what they learn and know, seeing what they see, and believing what they believe (not sure this is even possible). Basically, you have to disabuse your idea of complete objective reality and allow subjective , affective reality through your own experiences and emotions to inform and guide your perceptions and understanding of the nature of that person’s or group’s particular reality.   

Here’s a hypothetical example. Let’s say I’m curious to know more about a parent education program (this hypothetical example could apply to a phenomenon that takes place in an organization, village, community, etc.). Perhaps I have an evaluation client who wants me to conduct a study to determine whether the program’s processes and outcomes are what they were intended to be, or for no other reason than I’m curious about how effective such a program is in providing parents with knowledge and skills related to early childhood school readiness, how it might change or improve parenting practices, and how adequately parents are able to prepare their children for kindergarten as an entrée into the academic enterprise. There are two different ways of satisfying my curiosity. First, assume the reason for my interest is the latter. The problem, however, is that I don’t have a child between the ages of two and five - early childhood age - so I wouldn’t have any reason to take part in the program. Other than what I might read about the program, or anecdotal, informal evidence from talking to parents I might know who are in the parenting program, I wouldn’t have any idea about the kinds of things parents learn, the skills they develop, what they feel about themselves or their children as they go through the program, and how they use what they learn to prepare their children for kindergarten. If I didn’t know any parents and was overly ambitious I could scour the academic literature for studies published on parenting programs and read up on them a bit. Either way, I would know more about the program, but I probably wouldn’t understand it very well because I was relying on somebody else’s perspective or research rather than my own lived experience and all that entails (i.e., emotions felt and past experiences remembered). Second, if my interest was the former I would likely design and conduct a study, based on the questions my client had about the program, using qualitative (e.g., observations, interviews, focus groups, case studies, and content analysis of program documents), quantitative (e.g., validated, scaled measures; meta-analysis; and surveys/questionnaires), or mixed methods (i.e., quantitative and qualitative methods in a predetermined sequence) to determine the effectiveness of the parenting program’s processes and outcomes compared to a theoretical standard. After data was collected statistical analyses might indicate significant changes in parent knowledge and skills had I used some sort of pre- and post-measure or repeated measures design, and my sample size was large enough to provide sufficient power to detect an effect if one was present. With the right type and amount of data I might even have created a prediction model (s) to determine whether, and the extent to which, parents would likely use their newly acquired knowledge and skills to adequately prepare their children for kindergarten, and whether and the extent to which the children would be ready for kindergarten based on any number of indicators. Given the findings from my study I might again be able to say, yes, I know more about the program than I did prior to conducting the study, but can I definitively say whether I really understand the program without having experienced it as a participant and allowing my own experiences and emotions to add depth or color to that experience?    

In both cases there seems to be a clear discrepancy between knowing the program and understanding the program. Arguably, knowing the program from the perspective of program participants through observation, survey or interview isn’t the same as understanding the program from one’s own experience and perspective. Notably, for a study to be empirical there has to be both scientifically-based objective observations and subjective sensory experience. Usually, when we in a social science research methods-driven field discuss empiricism, we only discuss it from the scientific perspective thereby rejecting the subjective perspective given its inherent bias. This begs the question: can rigorous, objective research methods and subjective, affective, ethnographic experiences such as those promoted by Thompson be combined during an evaluation or other kinds of research to both know and better understand the nature of a program’s reality? The answer and premise of this piece is, yes, yes they can, the combination of which I’ve termed Gonzo Evaluation in homage to Thompson’s Gonzo Journalism. 

Combining objective and subjective methods may not work in every research situation or milieu for obvious reasons (e.g., research questions and study design don’t necessitate both approaches; no one to assist with entrée into the community, group, village, program, or organization; the required permissions weren’t granted; there isn’t enough time or budget to be both a participant and researcher; the sequencing of objective methods followed by subjective methods can’t be done; etc.). For situations in which it’s feasible, not only would we be able to know, objectively (i.e., surveys, scales, interviews, etc.), whether the program was implemented with fidelity to a theoretical model or whether outcomes were as hypothesized (determining impact requires a longer research time-line), we would better understand the program affectively from the perspective of what I call the participant researcher (akin to the ethnographer), the combination of which is what true empiricism, as previously noted, is all about. To be clear, the kind of participation I’m talking about is a different type of participation than in community-based participatory research (CBPR) or other forms of collaborative approaches such as empowerment evaluation (EE). In those approaches, members of a community, village, program, or organization actually develop a relationship with and perform evaluative/research tasks alongside the researcher as a way of building their evaluation or research capacity, skills, knowledge, and self-efficacy towards increasing a sense of individual and collective empowerment and self-determination. In Gonzo Evaluation the participant researcher is, for one component of the evaluation or research, embedded (think war correspondent) in the program, organization, community, village, etc. and does what a program participant, member of the organization, or denizen of a community or village typically does, in essence sharing their lived experience and making it one’s own. The distinction between using qualitative research methods such as observing from a distance and being a participant researcher is the proximity to the action, and the visceral, raw senses and emotions one experiences. However, getting in the game, so to speak, is only one part of the research equation in Gonzo Evaluation. While being a participant researcher is the subjective, experiential, affective, and ethnographic aspect of conducting this kind of research, Gonzo Evaluation also requires an objective component using any number of quantitative, qualitative, or mixed method approaches.  Simply stated, Gonzo Evaluation requires both an experiential, subjective, affective component as a form of ethnography, and a rigorous, objective measures observational component.

A little about the threat of bias (a.k.a., invalidities) in Gonzo Evaluation is in order. As a graduate student I was taught that if the researcher isn’t removed from the research environment or research subjects by, for example physical distance, emotional proximity, or personal identification to the greatest extent possible when running focus groups, conducting interviews, implementing surveys, or making observations then there is a strong likelihood bias will be introduced; anyone we’re observing, interviewing, surveying, or conducting focus groups with might act or respond to questions differently in our presence than they would otherwise in our absence. In addition, one of the first things an erstwhile graduate student learns is to recognize and account for one’s own biases when conducting research and evaluation. This doesn’t totally eliminate the threat of bias, but at least it’s recognized and ostensibly dealt with through continued vigilance as a form of checks and balances. Bias is bad and something to be avoided at all costs because once identified it calls into question the validity of the findings. That said, are there potential bias problems in Gonzo Evaluation apart from the bias problems inherent in the social science research enterprise? Indeed, I see two potential sources of bias: 1) if the subjective, affective experiences of the participant researcher accrete into the objective aspects of the research such that the researcher’s feelings, emotions, and experiences influence, for example, what is observed and how it’s observed, or what and how interview questions are asked (document content analysis or other forms of non-participant research notwithstanding); and 2), if the knowledge and insights about a program, organization, community, or village from having, for example, surveyed, observed or interviewed participants or members influences the natural behaviors and affective reactions one would have as a participant (e.g., behaving in ways that are expected of program participants or community members rather than how one would behave without those expectations within the same context). Unfortunately, bias works both ways in Gonzo Evaluation. In theory, however, the unbiased, objective, and rigorous component is supposed to serve as a counterbalance to the subjective, affective component. So, how do we handle potential invalidities in Gonzo Evaluation? I think I have a solution. 

As noted, there are two primary components in Gonzo Evaluation, objective measurements and subjective participation, so it’s a matter of timing and sequencing each component appropriately as well as choice of objective methods. If the evaluation is designed in such a way that the objective measures are implemented first and the subjective component is implemented subsequently then no matter what happens experientially the researcher/evaluator always has the objective measures to explain, or know, the phenomenon under study. Further, if only those objective methods are used that prevent the researcher from observing a program, organization, community, or village, or engaging in any way with participants or residents, as during an interview or focus group, then data collected through surveys and scales, perhaps even online, can be implemented and set aside for analysis after the researcher goes through the program or takes part in the affairs of a community; little if anything will be known until the analysis is complete. The whole point of Gonzo Evaluation is adding deeper understanding and insight to what one can reasonably know about a phenomenon through objective measures so it makes sense to observe first and try to understand second. So to minimize bias, first use objective measures that include either scales, surveys or both prior to participation. Only after participating would the researcher write up some sort of field notes or a case study with the addition of the affective aspects of the experience. Completing this exercise afterwards will keep the focus on participation and not on laboriously taking notes. There’s really no right or wrong in terms of what a researcher might write in a journal as subjectivity is personal. What one experiences as a participant can of course be influenced by previous experiences outside of the research or evaluation setting, but what the researcher feels while participating will likely be an accurate representation at that moment in time. Participating in a program or the affairs of a community isn’t a one shot deal so whatever biases there might be will be balanced out over time such that the affective response feels accurate and authentic. 

Getting back to the parenting program example knowing more about Gonzo Evaluation than when we started this adventure, for the objective part I would likely use both surveys and scales to determine parenting knowledge and skills related to preparing children for kindergarten and administer them before, during, and after the program (the same can be done with community or village residents and organizational members if a particular event or phenomenon is being studied). One of the assumptions of Gonzo Evaluation, and the only way I see it working unless someone comes up with a better idea, is that there has to be multiple sessions of the program, or the phenomenon under study is one sufficiently long in duration. If not, then relying only on one’s quantitative, qualitative, or mixed methods toolkit would certainly suffice for the sake of knowing, there would simply be no subjective component for the sake of understanding. As I said earlier, Gonzo Evaluation won’t work in every situation. I think it would be challenging to use objective methods to collect data while taking part in the program or in the affairs of a community or organization. I suppose if you were pressed for time and had budgetary constraints (when is this not the case in research or evaluation?) then you could do both simultaneously, but the program participants or community members might view you as a researcher first and participant second and likely not engage with you in an authentic way during that component of the evaluation. A trade-off might be to use pre-measures well in advance of when the program begins, before the researcher participates, skip the second measure altogether, and then conduct a post-measure at a sufficient period of time after the program has ended. When the participation period is over then it’s a matter of analysing the data from the survey or scale and using whatever qualitative analytical techniques are appropriate for the research or evaluation question (s) to analyze the subjective data collected in the journal. Interpreting the results of the quantitative analysis would follow the usual course of events, but interpreting the results of the qualitative analysis might prove more challenging given their affective and nuanced flavor, likewise putting the findings of both analyses together in some sort of report, data dashboard, or data visualization, however one wants to present said findings; it is not the intent of this piece to speculate on the nature of data analytical techniques, the interpretation of findings, or how to integrate objective and subjective findings into a cohesive and coherent whole. Rather, the key is blending objective knowledge about the program or community with that which is understood about the program or community to form a more complete picture of the nature of the reality of that program or community. In simple terms, the researcher would add personal perspective to the quantitative data as a way of explaining those findings. So, in the final analysis I can say that I both know and understand the program from having used both objective measures and from subjective experience.

Obviously, Gonzo Evaluation is not without its challenges, and doesn’t fit every research or evaluation situation. If, however, it fits the situation and challenges are overcome in a way that minimizes bias, constraints using both objective and subjective components in the same study notwithstanding, then I think the benefits far outweigh the costs and merits consideration as a new, hybrid type of research and evaluation that brings together tried and true objective research methods with an ethnographical component. Gonzo Evaluation, I think Hunter S. Thompson would be both pleased and amused. Thank you for reading. If you enjoyed this piece please read my other articles and leave comments as you deem appropriate.                 

References

Dewan, M. (2018). Understanding ethnography: An ‘exotic’ ethnographer’s perspective. Asian Qualitative Research in Tourism, 185 – 203. 

Plimpton, G. (1965). Paper lion: Confessions of a last string quarterback, New York, Little Brown and Company.

Thompson, Hunter S. (1966). Hell's Angels: The Strange and Terrible Saga of the Outlaw Motorcycle Gangs. New York: Random House.