"Epistemic Certainty Surrounding Dietary Recommendations for Meat"
Ellen M. StreetEllen Street is a Ph.D. student in Nutrition at Oregon State University. Her graduate research is
Contents |
DiscussionOverall Trends and Comparisons The observed dearth in booster usage among scientific writing genres may be best explained by Hyland; factual statements considered reliable, in scientific writing, are expressed using the indicative present (1996). If boosters are used, they are employed to express certainty explicitly (Pastore & Dellantonio, 2016). PHO publications used the highest number of boosters, followed by research articles. The DGA did not use any. Boosters are used in PHO and research articles to express certainty mainly pertaining to associations between processed meats and colorectal cancer, though there are interesting combinations of boosters and hedges to describe associations linking unprocessed red meats and cancer. Hedges Hedging is primarily used in scientific and government publications, motivated by a reluctance to commit to definite interpretations of the evidence. Markers of hedging are used to offer readers a wider array of interpretation, to indicate commitment to accuracy based on study designs or potential repercussions for false statements later on. PHO recommendations use the lowest number of hedges, in exchange for the higher modality imparted by directives and boosters. Government The current rhetoric in the 2015-2020 Dietary Guidelines for Americans reminds readers that the eating pattern itself influences health, rather than a single, isolated component, thus removing agency from red or processed meats as dietary risk factors. The statements in these guidelines lack epistemic certainty, as evidenced by lack of boosters and use of modal verbs may and can, expressing low and medium degrees of modality. There is evidence to suggest the recent DGA language shift and stance towards meat consumption might be influenced by submitted commentary from the NAMI. Alternatively, the language and word choice may have been intentionally implemented to avoid commitment to certain recommendations or conclusions, or used to create a more flexibility in nutritional recommendations to serve a wide variety of lifestyles and dietary needs. Public Health Organizations PHO publications chosen for this analysis employ significantly more boosters, directives, and less hedges compared to the other two genres, leading to a higher overall level of certainty. PHO recommendations regarding red meat are declarative and emphatic in their language, calling for limited red meat intake due to documented health risks and concerns at the community and population levels, such as cardiovascular related illnesses, and colorectal cancer. The confident tone imparted by directives and boosters used in these documents communicates definitive authority over health recommendations and reflects an ethical and humane imperative. Scientific Research Articles Markers indicating decreased certainty are more common among research publications, elucidated by Hyland’s rationale of hedging as an avenue for ensuring accuracy and precision in a field of organized skepticism (1996b). This is apparent in the Lancet monograph, AIM, and AJE articles. Although associations may be described with higher modality, the evidence is regarded with decreased certainty, presenting an interesting dichotomy[1]. The hesitance in commitment is understandable within this field, given that nutritional recommendations are constantly shifting and changing as knowledge and research techniques evolve. Interpretations and Discoveries Dietary Guidelines for Americans 2015-2020 The only booster in the 2015-2020 DGA describes the strength of evidence linking red meat, processed meats and poultry to CVD risk. Overall lack of boosters describing red and processed meat intake in the 2015-2020 DGA may indicate a lack of certainty in government recommendations regarding red and processed meat consumption and could express an unwillingness to explicitly support meat consumption or commit to specific recommendations. Perhaps limiting boosters describing red meat intake was intentional, allowing readers to draw their own conclusions. The choice to phrase the dietary guidelines as suggestions rather than explicit recommendations may also explain the absence of boosters. Over the course of my analysis, I discovered commentary, submitted by the NAMI, which challenged the drafted recommendations, arguing nutritional benefits and relevance of meat consumption. Additionally, the AMI offered that the DGAC phrase the guidelines as suggestions. The AMI argued for guidance, rather than prescriptive advice, to help people make healthy diet choices. The AMI concluded with a call for practicality, affordability, and attainability in dietary guidance to help people “eat a more healthful diet within the context of their existing food choices” (Backus, 2014b). Word choice in the recent DGA also reflects the July 2014 AMI commentary, particularly with hedge use. In the section titled “Meat, Poultry and Processed Meats Fit In A Balanced Diet,” AMI commentary authors Betsy Booren and Susan Backus use the phrases “can play a role in a healthy balanced diet” and “... processed meats can and do fit in a healthy eating pattern,” as long as consumers make choices that “fit within recommended nutrient and calorie needs” (Backus, 2014b). The discovery of hedging similarities between submissions from competing interests and the resulting recent edition of the DGA is important to note as the AMI commentary may have strongly influenced the recent DGA recommendations. Past editions of the DGA do not include full-fat or processed meats as part of a healthy eating pattern and explicitly recommend consuming lean meat as part of a healthy eating pattern, with explanations to clarify. Whether from outside influence or appeal to popular demand for dietary guidance over dictatorial constraints, the recent DGA guidelines show a marked shift from previous years in the level of epistemic certainty describing red and processed meat consumption. There may be multiple reasons for the inclusion of processed and full-fat lean meats in the recent DGA, but the main difference between editions published prior to the 2015-2020 recommendations is the shift in language surrounding this red and processed meat consumption. American Institute for Cancer Research – “Food, Nutrition, Physical Activity and the Prevention of Cancer: a Global Perspective – Recommendations for Cancer Prevention” The WCRF/AICR 2nd expert report contains boosters specifically describing the relationship between processed red meat and colorectal cancer, communicating high levels of epistemic certainty. This reflects the breadth of evidence supporting associations between colorectal cancer and processed meat with boosters “stronger” and “even more clear-cut” describing the evidence. Contrasted with the language used in the 2015-2020 DGA, the difference in certainty regarding processed meat consumption is drastic. However, the relationship between red and processed meats to other types of cancer besides colorectal is regarded with much lower certainty, evidenced by low modality markers and hedges. Plausibly, an accuracy oriented modal adverb, indicates a low level of epistemic certainty and probable, a reliability hedge, indicates the writer’s confidence in the truth of the statement (Hyland, 1996b), yet fails to concretely assign agency to processed meats as a cause of stomach cancer. The differences in certainty between the purported cause and effect of red meat consumption is intriguing. The antithetical description of red meat as a “convincing” cause of colorectal cancer (see appendix table 5) conflicts with the hedge used to describe the mechanisms by which red meat may cause cancer. Clearly, higher certainty lies with processed meat as carcinogenic, but the article leaves room for red meat to be considered carcinogenic as well. AHA – “Recommended Dietary Pattern to Achieve Adherence to the American Heart Association/American College of Cardiology (AHA/ACC) Guidelines: A Scientific Statement from the American Heart Association” The guidelines do not use boosters or hedges in descriptions of red and processed meat intake. Instead of boosters, these recommendations use directives to authoritatively instruct their readers on how to follow a heart-healthy diet. The choice to use directives instead of boosters is interesting, particularly because of the strength of certainty it implies. Even more so, the reference to the 2015-2020 Dietary Guidelines Advisory Committee “[who] summarized data from 55 studies" and concluded that there was strong evidence that any of the above eating patterns [MED or DASH]...are “lower in SFA, cholesterol, and sodium and higher in fiber, potassium, and unsaturated fats are beneficial for reducing cardiovascular disease risk.” This statement corroborates the AHA directives for readers to "limit intake" of red meats and "reduce sources" of meat and dairy fat to lower trans-saturated fat intake. The use of these rhetorical markers communicates a definite authoritative stance towards recommended dietary restrictions of red meat. The most interesting aspect of this reference is how it relates to the 2015-2020 DGA red and processed meat recommendations. The AHA interpret the MED as an appropriate example of a low saturated fat, heart healthy diet, repeatedly shown to be effective at decreasing CVD risk. It’s ironic that the DGA cited this information, yet fails to discuss the merits of this eating pattern in the sections discussing red and processed meat intake. The difference in target audience for the AHA recommendations may explain the use of directives rather than boosters. While the DGA is directed towards the general public, the AHA recommendations are customized for individuals with risk factors for stroke and heart disease. The authors of the AHA recommendations may be using authoritative, clear-cut guidelines to reach these targeted individuals with a higher mortality risk who need a strong, clear message to motivate changes in dietary and lifestyle patterns. Diet continues to be the cornerstone of care for individuals with metabolic disease and I believe that the absence of hedging may be intentional, in order to communicate necessary dietary advice. The Lancet Oncology – “Carcinogenicity of consumption of red and processed meat”: In Ken Hyland’s article "Genre and Academic Writing in the Disciplines," he describes the spread of knowledge in the sciences as acceptance of new knowledge through experimental proof (2008). The strength of the experimental proof bolsters the acceptance of new knowledge. High modality markers and a lack of hedging indicate certainty and acceptance of the connection between processed meat and colorectal cancer risk, while lower modality markers describing red meat and colorectal cancer risk point to an uncertain association. Though there are boosters used to describe the link between red meat to colorectal cancer, the overall level of modality describing red meat’s link to pancreatic and prostate cancer is lower than the modality supporting the link between processed meats and colorectal cancer.[1] The language and conclusions in the Lancet monograph reflect those drawn in the AICR/WCRF 2nd expert report, published eight years prior. The Working Group uses probably to describe associations between red meat and cancer, indicating a low modality, followed by boosters (“substantial”) describing the supporting evidence for the classification of red meat as “probably” carcinogenic. “Substantial” indicates a significant body of evidence to support the Working Group’s hedged claim, while “sufficient,” used to describe the classification of processed meat as carcinogenic, indicates only an adequate amount of certainty. The descriptions of the evidence do not match up with the indicated levels of confidence behind the evidence.[2] This may be interpreted as avoidance of personal commitment, in case new evidence proves their classifications erroneous. This is a reasonable given the tendency of nutritional recommendations to change over time. Archives of Internal Medicine – “Red Meat Consumption and Mortality” The sole booster in this article is used to describe an association between heme iron in red meat, MI, and CHD. The statement claims "…dietary iron, particularly heme iron primarily from red meat, has been positively associated with myocardial infarction and fatal coronary heart disease." Though this is not a strong booster, it does indicate a higher level of certainty over a standard association by indicating a stronger source of evidence than if the association lacked a descriptive adverb. This article contains a higher proportion of low modality markers used when referring to constituents of processed and unprocessed red meats shown to be associated with increased CVD and cancer risk (see appendix table 7). Additives and byproducts of cooking; nitrites, nitrates, sodium in processed meats, heme iron, and heterocyclic amines, are given agency as carcinogens, rather than the meat itself. The markers used to describe specific deleterious components of red and processed meats (may, might, potential) are reliability hedges, based on current information about the additives and additional components comprising processed meats. Interestingly, the conclusion indicates higher certainty for mortality risk from red meat consumption while the discussion expresses hesitance using reliability hedges to describe the harmful health outcomes of meat consumption. This may be explained by the team’s examination of prospective studies, and like the working group, hedging could be used here to avoid personal commitment to interpretations. American Journal of Epidemiology – “Red meat and processed meat consumption and all-cause mortality: A meta-analysis” Larsson and Orsini use boosters to describe the connection between processed meats and colorectal cancer, while utilizing hedges to link red meat consumption to deleterious health consequences. These statements mirror the attitudes exhibited in the Lancet monograph, the WCRF/AICR expert report and the AIM prospective cohort study. The use of boosters also reflects similar usage in the WCRF/AICR and Lancet publications. The description of the association between red, processed meat, and all-cause mortality as positive, particularly in an epidemiological meta-analysis, indicates a stronger stance towards the evidence than it would if the association had not been paired with a preceding adjective at all, akin to its use in the AIM article above. Likely, the authors chose not to use a descriptor with higher modality due to the wide range of factors that could have influenced an association between meat-eating and all-cause mortality. The number of meat-eaters is too great to lend itself to parsing out credible associations. The use of most consistently likely references the plethora of evidence backing associations between processed meats and cancer, which may indicate a higher strength of evidence compared to that of red unprocessed meat and colorectal cancer. This indirectly boosts the significance of evidence linking colorectal cancer to red and processed meats without the authors claiming ownership over the relationship between the two. The authors use one reliability hedge, may, which allows room for other explanations and alternatives to the conclusions made in this paper, undermining the author’s “confidence in the truth of a proposition, as warranted by the available facts” (Hyland 1996b). This could have been used to show the author’s commitment to accuracy in their claims; after all, this publication was a meta-analysis and not a cohort study or randomized controlled trial. Perhaps the authors were reluctant to commit to certain interpretations linking all-cause mortality, chronic disease, and cancer to meat without stronger associations or evidence.
Topics for Future Research My research revealed several intriguing findings related to my topic of analysis. First, I noted the presence of manipulative silence (Huckin, 2002) employed by the DGA committee, likely as a response to the NAMI commentary. Also, I noticed differences in certainty expressed by the Working Group in the AICR/WCRF 2nd expert report. Connections between red and processed meats with carcinogenic outcomes are described with lower modality than the evidence supporting the classifications. There is also a stark divergence in certainty attached to associations between red meat and types of cancer besides colorectal in the Lancet monograph. Lastly, the AHA recommendations draw upon the Mediterranean diet, sometimes termed the MeDi or MED, as a “beneficial [diet] for reducing cardiovascular disease risk.” The AHA cited the 2015-2020 DGA for this information. Obviously the DGA is aware of eating patterns that reduce CVD risk, but fails to highlight the evidence in the sections where it would be best applied to recommendations. Together, these discrepancies and idiosyncrasies present fascinating opportunities for a closer look at the influence of outside forces on dietary guidelines for the U.S. population and attitudes surrounding red and processed meats as carcinogenic agents. Limitations My analysis of the selected materials revealed an interesting array of markers and ambiguities related, but outside the scope of this paper, that would be excellent topics for further research. This analysis was limited to analyzing epistemic certainty through booster and hedge use. Links between certainty and other markers such as stance, relevance markers, proximity and code glosses were not explored. Certainly, a closer, more extensive analysis would reveal a complex web of connection between certainty, concession, and emphasis. The focus of my analysis may have shifted due to my own interest in the subject, surrounding controversy over red meat consumption and political conflicts of interest I discovered throughout my research. I attempted to avoid bias by focusing on the rhetorical moves, hedges and boosters, supporting my statements with credible citations, and sourcing evidence to corroborate the patterns I found. ___________________________________________________________________ [1] While this ambiguity is important to note, it does not specifically relate to the goals of the paper. Further research is suggested. [2] The difference in meaning between these two adjectives and their respective related claims is intriguing, but beyond the scope of this analysis.
[1] And a fascinating opportunity for future research. |