The topic of factory farming has come to be a very complex matter. Involving a plethora of actors and producing profound social, economic, health and environmental consequences, the subject holds relevance in many fields of study. Thus, my research will combine critical analysis and discursive study in order to examine the issue of factory farms. Moreover, I intend for this research to shed some light on the available literature as well as the institutional actors at play. However, prior to the details, I will provide some general background knowledge to set the scope of this paper. Throughout the paper, the terms factory farms and CAFOs (Concentrated Animal Feeding Operations) will be used interchangeably. As such, The Cambridge Dictionary defines factory farming as, “a system of farming in which a lot of animals are kept in a small closed area, in order to produce a large amount of meat, eggs, or milk as cheaply as possible” (Factory Farming, n.d.). The advent of factory farms marked a prolific point in history due to the resulting production of massive quantities of food for a growing human population. However, it has enabled very costly ramifications for the environment and humans alike. Likewise, there are a variety of actors involved in this process, namely corporations, NGOs, governmental bodies, and lobbying groups. Therefore, I inquire, “What degree does the language of these actors reflect their true intentions? Moreover, what sentiment does each actor demonstrate in relation to CAFOs, and what other institutional factors may be at play?”
Of course, one must ask, “Why are these questions important in the first place?” They are important because language is a discernible marker of how actors attempt to portray themselves to the public. Likewise, asking these questions allows us to see how these major players view the issue (or non-issue) of factory farming and its subsequent effects. Ultimately, these questions are significant because their responses reveal the true extent of the alignment between the actors’ underlying intentions and their public story. We must have a way to evaluate this relationship, which is precisely why I implement sentiment analysis, an objective form of data inquiry. This is done with the hope that we can highlight the institutional factors at play and their role in policy-making. This is a multifaceted issue, and my analysis attempts to bring various factors together. Ultimately, discourse and policy are intimately related, which makes language a powerful analytical tool. Thus, I will begin my analysis with a formal literature review and follow this with the research itself.
As the prevalence of animal agriculture has increased in modern times, it has severely exacerbated the effects of global climate change. The primary evidence for this impact is through greenhouse gas emissions. Cole et al. (2009) note that, “meat and dairy products contribute more than 50% of the total GHGs emitted from the food chain,” (p. 162). Meanwhile, the FAO of the United Nations reports a figure of “18% as the global contribution of the animal farming sector to GHG emissions (Steinfeld et al., 2006)” (Cole et al., 2009, p. 162). While this is not as significant as the energy sector, it marks an important contribution to the global climate. This is precisely why Garnett (2009) says, “[M]eat and dairy products are the foods carrying the greatest environmental burden” (p. 491). Moreover, this problem has proliferated at an accelerated rate because of the very structure of the capitalist system. Mass-scale efficiency is prioritized over smaller, sustainable modes of production. While the United States enjoys unprecedented outputs of food, it risks environmental calamity as climate change continues to worsen. GHG emissions are far from being curtailed, one study reporting that “[global] emissions could increase by a further 17 percent by 2050 (Bennetzen et al., 2016)” (Garnett, 2017, p. 33). Furthermore, accelerated climate change is not the sole consequence of factory farming.
With the increased demand for meat output, the animals themselves require increased input. This has resulted in detrimental environmental impact, as the farmed animals (specifically cattle) demand large quantities of soy and grain for feeding. Unsustainable practices such as the continuous production of corn and soy cause the land to become arid (Garnett, 2009, p. 493-494). Likewise, this endless production has caused vast swathes of forest land to be cleared (Garnett, 2009, p. 494). This is done on a massive scale, all for the raising and production of livestock (Garnett, 2017, p. 39). Coupled with overgrazing, this has caused desertification, which has effectively reduced the amount of usable land (Horrigan et al, 2002, p, 447). Animal agriculture has indirect impacts on water quality as well. Cassuto (2010) explains, “Worldwide, such livestock-related land use changes as deforestation and overgrazing cause desertification, ground water and soil contamination, and other environmental problems” (p. 9). CAFOs cause a myriad of environmental harms, each problem linked to another, such as sea level rise, increased natural disasters and loss of biodiversity. Of course, factory farms are only exacerbating these impacts. This is precisely why the Millennium Ecosystem Assessment (2005) called factory farming the “largest threat to biodiversity and ecosystem function of any single human activity” (as cited in Weis, 2010b, p. 136). Overall, the ecological footprint of CAFOs is potentially devastating if it continues at such massive rates. If we wish to prevent the worst of the consequences, we have to ask, “How has this problem manifested in the first place?” I argue that we must examine the nature of capitalism in order to answer this question.
Now that the link between factory farming, environmental degradation, and climate change has been established, I address the economic foundation that allows for it. Capitalism has achieved something amazing in that it revolutionized the means of production, so much so that we have evolved into a massive consumer society. Now more than ever, we can get anything we need with the upmost rapidity. However, this comes with a dire cost; there are severe biophysical consequences for such increased production (Weis, 2010a). This massive level of consumption requires an equally large amount of input, namely through natural resources. Thus, the capitalist class accumulates capital by exploiting the environment for its resources. A quintessential example of this exploitation is the burning of fossil fuels to produce energy. As Weis (2010b) points out, “The failure to account for the atmospheric burden associated with fossil energy, and its impact on the Earth’s climate system, represents one of the most fundamental biophysical contradictions of industrial capitalism” (p. 318-319). This contradiction is a fundamental aspect of our current climate crisis. There is an inherent conflict between the way in which the capitalist system produces mass goods for a consumer society and the way in which capital overuses the environment (Moore, 2016). Interestingly, this is why Moore (2016) claims that the world is not in the age of the Anthropocene, but the “Capitalocene.” He explains that world history is in “the ‘Age of Capital’—and the era of capitalism as a world-ecology of power, capital, and nature” (p. 6). Therefore, capitalism pursues its power through the co-production of nature by transforming it into a commodity. Although, it seems that the environmental consequences of production will eventually prevent the further accumulation of capital because these ramifications will create a limited supply of resources and fundamentally alter Earth’s climate system, thereby making it harder to produce goods.
Thus, we have found ourselves in a scenario akin to the tragedy of the commons. As corporations take natural resources or pollute the Earth so much so that we can no longer obtain resources, the system will hit an inescapable point of collapse. In theory, capitalism provides the best allocation of resources which in turn produces economic growth and social prosperity (Magdoff, 2015). However, the system does not function this way—capitalism has overexploited and endangered its very foundation of power. The Age of Capital relies on nature for its success, but if capital accumulation relies on the unsustainable depletion of natural resources, capitalism runs on a damning contradiction (Magdoff, 2015). The consequences for both the Earth and the future accumulation of wealth are disastrous. Referring to the tragedy of the commons, Magdoff (2015) states, “A rational economic decision for each individual farmer goes against the supposed capitalist economic logic and ends up being irrational for the entire group of farmers together” (p. 13). Similarly, as Weis (2010a) argues, all of the rational capitalists have overexploited in order to accumulate massive amounts of capital. Unfortunately, this has resulted in an irrational outcome for the entire group. Weis (2010a) states, “[T]he deceptive efficiency of industrial capitalist agriculture and its manifestation in cheap, bountiful food have long overshadowed the instability and inequalities of the system” (p. 317-318). So, the question that we must inevitably ask is this: “Can Capitalism solve a crisis of its own making?”
Unfortunately, the answer that I provide is not very optimistic. I say this because the United States’ excessive consumption and increased reliance on factory farms may make a capitalistic solution impossible. High rates of consumption contribute to growing levels of greenhouse gas emissions from the animal agriculture industry (Cole et al., 2009). Thus, I must analyze the issue of animal agriculture in relationship to the greater issue of mass consumerism.
To begin, Lavin (2009) argues that industrial animal agriculture rose to prominence as it integrated “the mechanization, division of labor, and intensified regulation of activities endemic to industrial manufacture” (p. 72). More so, the advent of factory farming came with the increased demand for meat and dairy while capitalist production continued to accelerate. In this way, farming became a factory system because its sole purpose was to exponentiate the production of goods for consumers (Lavin, 2009). Moreover, Lavin marks a second driving factor of consumption: capital concentration (p. 73). While the factorization of food production certainly sounds like an appropriate solution for a growing population, the incentive to factorize has been driven by the desire for more wealth. Lavin makes a compelling point as he analyzes various critiques of factory farming. Namely, “the common theme of these critiques, similar to others of industrial life, is that the pursuit of efficiency causes great harm to people, animals and the environment” (p. 76). Therefore, industrial life does not just produce negative outcomes for the animals, but also for the environment and the humans in it.
Perhaps the most important factor when it comes to the rise of animal agriculture in the age of consumption is the commodification of animals. Capitalism has transformed living beings into products that are bought and sold on the market. This fact is crucial to understanding the United States (and global) food system because it reveals the way in which capital can control what the public eats. As such, Gunderson (2013) explains that “the primary purpose of rearing livestock today is not to create food, but to make money – or, for exchange values, not use values” (p. 261). This is precisely why authors have been raising awareness about the possible health concerns associated with industrial animal agriculture (Horrigan et al., 2002). Production for exchange value is not a new thing, but rather, a historical fact of production that has emerged in tandem with mass consumerism. Likewise, this is not unique to factory farming. In fact, capitalism has relied on the categorization of exchange value as its basis for capital accumulation. Unsurprisingly, Marx pointed this out long ago: “Use-value, not exchange-value, is the purpose of the whole system of production, and use-values accordingly cease to be use-values and become means of exchange, or commodities, only when a larger amount of them has been produced than is required for consumption” (as cited in Gunderson, 2013, p. 261). Evidently, mass consumerism is only a new stage in the process of capital accumulation.
The process of commodification has accelerated over the years, resulting in unprecedented levels of meat and dairy consumption. As Lavin (2009) demonstrates, the power of food production now lies in the hands of corporations. Not only does this have ramifications for the animals themselves, but also for the consumers. Lavin explains how in becoming a consumer society, the burden of discipline now falls onto the consumers instead of producers (p. 85). It is no longer the obligation of corporations to engage in ethical production; this notion comes with a plethora of quandaries and injustices. This newest stage of commodification enables corporations and producers to move the blame to consumers. Meanwhile, Lavin argues that this has marked a transformation in the way corporations control their workers and the public. Quoting two prolific authors, Hardt and Negri, Lavin explains, “Now, individuals are monitored and regulated not only when they visit the factory, schoolhouse, or hospital, but also in more prosaic and ubiquitous ways in their daily activities; disciplinary power spreads across this space, with the result that coercive authority organizes subjects ‘in the totality of their activities’” (p. 84). Hardt and Negri demonstrate how corporations intervene by accumulating information on us. This is no surprise in the age of big data. However, the individual’s information is now collected in a more casual manner, through the products one consumes. It is this buying history that informs corporations about each person in society.
Furthermore, Lavin (2009) goes on to show the intense effects of mass consumerism. He states, “My position in cultural politics depends not on my status as worker or owner, but my preference for NASCAR or lattes. This is surely why the factory metaphor travels to so many domains of life: the concerns about alienation and control that inhere in the metaphor are not anachronistic in an age of casualized labor; they are instead generalized to work, leisure, and politics” (p. 85). In this sense, commodification has bled into all aspects of society, including politics. Corporations can now control consumer life through indirect and untraceable means. Likewise, commodification is intimately connected to the discourse corporations use to attract consumers. Under modern capitalism, this is the only means of control that corporations need. Corporations dominate the consumer narrative, selling people the idea that they need to eat more meat. They shape discourse in a way that appears to shift decision-making onto the consumers, while masking their true intentions. Gunderson (2013) explains how the notion of ethical consumerism is simply a means to defer decision-making to the individual without effecting significant change. “‘Ethical consumerism’ is a reflection of bourgeois market ideology that alienated individuals are expected to subscribe to,” he says. “The problem is that individualist ethical consumerism is not only limited and ineffective in the face of larger socioeconomic forces, but it also halts social justice movements from pursuing radical means of altering society because they have been co-opted” (p. 269). The author makes it incredibly clear that not only is the promotion of ethical consumerism a calculated move by capitalist powers, but it is also an extremely ineffective tool when it comes to creating meaningful change. Instead of trying to change the system that allows this, ethical consumerism reduces actions to the individual level. As we will later find out, this is reinforced by a manipulative rhetoric, furnished by corporations and rich investors to further their agenda.
Of course, many still argue that capitalism is not the cause of this crisis. Green investors and the like try to argue that a free market solution is just beyond the horizon. Corporations have now “greened” themselves by transitioning to more sustainable energy usage and more environmentally friendly practices. However, authors have rightly pointed out that these individuals are simply trying to capitalize on the downfall of fossil fuels. While the capitalists may be taking measures that seemingly benefit everyone, they are only attempting to gain wealth for themselves. Regardless of their prerogative, capitalism’s goal is endless growth, and that can never be compatible with sustainability (Guerrero, 2018). Likewise, Big Green (highly influential environmental organizations like the NRDC, Greenpeace, Nature Conservancy etc.) are not a great option either. All of these entities attempt to address symptoms of the climate crisis rather than the root of the problem: capitalism. Furthermore, Naomi Klein (2014) claims that these Big Green organizations are actually linked to fossil fuel companies. They attempt to implement policies that are seemingly better for the environment; meanwhile, they are sympathetic towards corporations (Klein, 2014, p. 198). This is no surprise considering Big Green has often taken funding from fossil fuel corporations. Klein goes on to explain that “several of these groups have consistently, and aggressively, pushed responses to climate change that are the least burdensome, and often directly beneficial, to the largest greenhouse gas emitters on the planet” (p. 198-199). Thus, environmental NGOs are attempting to institute a green capitalism that will only result in austerity and further environmental degradation.
This form of “sustainable capitalism” is an oxymoron in itself. Guerrero (2018) argues, “A growth-driven and market-dependent system is incompatible with environmental security” (p. 43). In actuality, corporations simply try to disguise their actions as green and eco-efficient so they can build public support while continuing their exploitative actions. For this reason, Magdoff and Foster (2010) assert that “a system that has only one goal, the maximization of profits, has no soul, can never have a soul, can never be green, and, by its very nature, it must manipulate and fabricate whims and wants” (p. 19). This two-faced nature of capitalism is seen in corporate rhetoric which disguises its sentiment towards factory farms as positive. The truly unsettling part is that many of these claims are nothing but ploys, just a means of masking the unethical exploitation. In fact, Clark (2012) explains, “[W]hat is perhaps most troubling of all about these practices is that an even shrewder, more calculating, and more ruthless efficiency is being celebrated as ‘ecoefficiency,’ greenwashing the underlying violence” (p. 120). This is in complete accordance with capitalism’s nature; it has always tried to hide its unjust and unethical practices. The oppression of animals and workers alike is simply a byproduct of capitalism. Just as I pointed out previously, the heart of this problem lies in the inherent contradiction between capital and the environment. This being said, the most significant aspect of green capitalism is not its contradiction, but the façade it creates. Capitalism has managed to co-opt discourse and maintain power by manipulating what the public hears.
While actions are a key indicator of how an actor regards a particular issue, discourse can be much more useful in showing how an underlying sentiment may define those actions. Therefore, I spend a considerable amount of time detailing the actions of factory farm stakeholders and highlighting how these various actors’ decisions are intertwined with one another. This coupled with a discursive analysis can reveal the political alignments of the actors as they pertain to animal agriculture. In other words, I will reveal the underlying coalitions that form around the issue of factory farming and greater environmental problems such as climate change. Moreover, this will help explain how our system comes to make decisions and whether confounding factors (such as increased capital) disrupt ethical policymaking. This section will fulfill two important roles within my analysis: it will establish the significance of discourse in politics, and it will demonstrate how current authors interpret rhetoric surrounding issues of the environment and animal welfare.
Foremost, I want to provide a discursive basis for why discourse within politics is even worth examining in the first place. Therefore, I will employ elements of Michel Foucault’s approach to discourse as a means of showing discourse’s power to shape society. Freeman (2009) helps offer some insight into Foucault’s philosophy as she explains, “Foucault felt discourse serves to define the limited ways that were suitable to talk about, treat, or engage the topic so its meaning is comprehensible to a society” (p. 83). It is important to understand that this has powerful political implications, as institutional agents have the means to decide how this information is conveyed to the public. In many ways, agents such as corporations, governmental bodies and NGOs have the power the shape perception and opinion. Furthermore, these actors can work to formulate a dominant ideology which controls how we view an issue. In this way, discourse analysis can be a very useful tool for delineating what constitutes suitable engagement or rhetoric when it comes to factory farming. Likewise, Leipold et al. (2019) believe that discursive analysis is especially useful for understanding why things happen and what sort of political motivations lie behind them. So how does this apply to our endeavor? Well, the way in which society approaches animal production is determined by how institutional actors talk about the issue (Freeman, 2009, p. 84). In addition, the public often adopts the same sentiment as its institutions through the things it reads. Thus, discourse acts as an important identifier of both actor and public attitude towards structures such as factory farms.
In order to better establish the role of discourse in politics, Leipold et al. (2019) offer the following explanation: “[T]hese patterns also determine the understanding of specific practices and events in the policy process, they are fundamental to the formation and expression of political truth claims, the engagement in, as well as the self-positioning of individuals and collectives for or against policy change” (p. 447). This seems especially pertinent to the issue at hand, as a careful analysis of the rhetoric of these actors reveals the underlying patterns of their attitude. Even more important, the discourse they use determines what policies are implemented and whom the policy decisions favor. Leipold et al. go on to explain, “The contributions to this special issue demonstrate that discourses can affect change or inertia at various levels of policy making – from discourse itself and its rules to policy outcomes and institutional settings (p. 452). Of course, all of this is contextual, and the rhetoric of an individual actor is largely dictated by the perception they want to create. Regardless, discourse has a clearly discernable impact on policy and on the associated decision-making. The way society’s various institutions treat factory farming informs the status quo and any potential alternatives to the current state of politics.
Unfortunately, language can have some negative effects in terms of policymaking and its consequences. While this discursive analysis is meant to shed light on the types of sentiment expressed by these actors, I must bring up prior assumptions that can help form my hypotheses. Firstly, as Glenn (2004) recognizes, corporations, and the U.S. government for that matter, tend to treat animals as objects rather than living beings. As the author states, “It is a recognizable discursive move that removes the ‘beingness’ or subjectivity of animals and replaces it with a word that morphs a subject for its own purposes into an object for consumption” (p. 69). As I suggested earlier, a corporation will employ positive rhetoric to create a benevolent persona that truly cares about the animals it kills. Meanwhile, this may prevent the public from realizing the truth – in other words, positive sentiment makes it less likely that people will recognize the animals’ pain (Glenn, 2004). Of course, one would expect this from corporations as they attempt to maintain their positive image and continue their profit-making. However, the position of NGOs is much different from that of the government. While I would expect the government to try to remain neutral on this position as it tries to provide just the facts, environmental NGOs are much more likely to be critical of the issue. This difference is predominantly because each of these actors occupies a different position within institutional politics. Accordingly, each employs a unique discursive strategy – though I do not suggest that this investigation is completely black and white, as the actors blend into one another at times.
I will undergo this research with the aforementioned expectations in mind. There is no doubt that actors of different categories have competing interests, which leaves them in a clash of discourses and public effect. Each category provides a different interpretation of factory farming. This is largely because each comports itself differently with regard to the issue. Moreover, each occupies a different perspective and has a unique goal when it refers to factory farming. For example, a corporation likely perceives CAFOs as a means toward profit, something that must be kept according to current regulations. However, corporations also want to retain public support to create more business; thus, they try to win over the public through their discourse. Manipulation of discourse allows them to paint the issue in their own terms while gaining public support. Alternatively, the government has no such agenda; its discursive strategy is to state the facts and explain the topic in its apparent truth. In doing so, government agencies provide an objective position based on the evidence they have. What is key is that the government is not attempting to win over the public with its discourse. NGOs are an interesting case because they certainly have an agenda, but one that is significantly different from corporations. They may try convincing the public that the issue of factory farming is important and needs comprehensive reform. Therefore, they depict CAFOs in a negative light, possibly highlighting their environmental harm in order to make the public see CAFOs as negative. Consequently, this creates a clash of discourses as one agent tries to make factory farming seem positive, the other negative, while the government remains ambivalent. Thus, with these discursive elements in mind, I formulate my hypotheses as such:
- H1: Corporations and lobby groups will use more positive language in their depictions of factory farming.
- H2: Governmental bodies will offer a neutral point of view (balancing use of positive and negative language) in their depiction of factory farming.
- H3: NGOs will use more negative language in their depictions of factory farming.
In order to undergo this discursive analysis, I will analyze a total of 62 documents, produced and published publicly by the actors outlined below. I haven noted a range of actors that hold salient information pertaining to industrial animal agriculture. As such, I have divided the actors into four categories: corporations, governmental agencies, lobbying groups, and non-government environmental organizations. These categories are thought to represent the actors’ goals as they pertains to factory farming. I should note that all these actors reside in the United States. Accordingly, I have chosen several from each category:
Table 1.1 List of actors, by category, from which documents were taken for the purposes of this discursive analysis.
Furthermore, I have chosen these specific actors for differing reasons based on their category. The corporations were chosen based on their impact and levels of production in the animal agriculture industry. In short, these are the companies who produce the most meat and dairy for consumption in the United States (Sharma, 2018). Alternatively, the various government agencies were chosen based on their relation to the issue, the most pertinent being the FDA and USDA since they had the most available information. Similarly, there was a narrow range of lobbying groups that I could have used for this analysis, so I have chosen the few that have a measurable impact on Congressional lobbying. Finally, the environmental NGOs were chosen based on available literature related to factory farming as well as their discernable impact on U.S. politics.
In establishing the most relevant actors of this study, I will now move on to why I have chosen the outlined documents for my sentiment analysis. I should preface this by saying that I searched for specific buzz words that are related to the issue at hand, namely, words such as “factory farming,” “animal agriculture,” “climate change,” “industrial agriculture,” and “animal welfare.” For the purposes of this inquiry, I sifted through open access documents furnished by each actor on its respective website. In the case of the government agencies, I used their database of documents as it pertains to the issue of animal agriculture. The NGOs and corporations were rather different because they did not provide a list of their documents as PDFs. As such, I often resorted to referencing their webpages, as this provided a concrete (yet concise) depiction of their communications meant for the public. In each scenario, the documents detail the actor’s discourse as it approaches the issue of factory farming. Even at first glance, the actors’ discourses vary greatly according to their various institutional positions. Therefore, the following section outlines my approach to evaluating these texts through Nature language processing and text analysis.
To analyze the data, I use sentiment analysis in the coding language Python. Sentiment analysis is a branch of data science that uses natural language processing and text analysis to extract information from a given piece of text. This is predominantly used to categorize specific words in the text and assign them sentiment values which will ultimately determine the writer’s (or actor’s) attitude toward the topic. This method is incredibly useful for a variety of fields, and I believe it to be especially applicable to this one. Hutto and Gilbert (2014) provide insight on the cross-disciplinary nature of sentiment analysis, stating, “Sentiment analysis is useful to a wide range of problems that are of interest to human-computer interaction practitioners and researchers, as well as those from fields such as sociology, marketing and advertising, psychology, economics, and political science” (p. 216). Consequently, sentiment analysis is a very valuable tool to test my hypotheses because it will allow me to inspect documents from each individual actor and compare sentiment scores. Likewise, each of these actors is likely attempting to produce a distinct sentiment in its writing, which will become evident through this study. In undergoing this analysis, I will examine the discursive elements behind the actors’ language and consider what these elements mean for their attitudes toward factory farming.
I utilized Python with a variety of packages to facilitate this endeavor. However, the most important and integral of these is VADER (Valence Aware Dictionary for Sentiment Reasoning) (Hutto and Gilbert, 2014). This acts as the core sentiment identifier for this project by utilizing its lexicon to determine a positive to negative compound ratio. Hutto and Gilbert (2014) explain how they “use a combination of qualitative and quantitative methods to produce, and then empirically validate, a gold-standard sentiment lexicon that is especially attuned to microblog-like contexts” (p. 216). I should note that the VADER is a tool designed especially for social media contexts. However, there are three reasons why VADER is a sufficient mode of analysis for this project. Firstly, Hutto and Gilbert acknowledge that this engine can be used “across several domain contexts,” as this tool is extremely accurate in identifying sentiment regardless of the text’s origin (p. 216). Additionally, the authors note that VADER has been found to be even more accurate than individual human raters – the model has a 0.96 classification accuracy as opposed to 0.84 (p. 216). Finally, the focus content of this inquiry is very similar to that of the original VADER subject matter. These documents were produced for public eyes on an online domain, and the sentiment of each individual word is likely to remain consistent across platforms. With that being said, I will now provide a brief explication of the coding process.
I began by importing all the necessary packages, the most important being NLTK (Natural Language Toolkit), VADER, pandas, matplotlib and wordcloud. Next, I loaded the document (according to the category of actor) into Python and downloaded the VADER-specific lexicon. Then, I implemented the basic steps of completing sentiment analysis, in the following order: I tokenized the data (splitting it into individual words), put all words into lower case, stripped the text of punctuation, and eliminated all stop words. Next, I lemmatized the data, a process which removes plural forms and groups like words together (e.g. “great” and “greater”). Subsequently, I took the lemmatized text and removed characters that were not needed, such as numbers and any other forms of punctuation that may have slipped by. Finally, I created an Excel sheet to organize the list of sentimented words with their scores into a data frame using pandas. From here, I was able to obtain the list of sentimented words with their corresponding compound score as well as information about the dataset such as the mean and standard deviation. My final steps correspond to the way in which the data has been visualized, which includes an assortment of graphs and word clouds. For more information, the entire script is listed below in Appendix A.
From this analysis, we have gained several insights into the sentiment value of the cleaned words for each of the four categories. This includes the means, along with other notable data features as displayed in Table 2.1. The average value is the combined sentiment score of all sentimented words, which ranges from a score of -1 to 1. These words and their values were determined by the VADER lexicon and are further detailed in Hutto and Gilbert’s paper. As rendered below, one can see that all of the sentiment averages are on the positive side, ranging from 6.8% to 24%. This indicates that some types of actors were attempting to create more positive images in their rhetoric than others. Of course, corporations and lobbyists have the highest positive mean scores, which coincides with Hypothesis One. Likewise, the government discourse was positive but very close to 0, indicating that it has taken a near neutral stance with a slight lean toward positive. This supports Hypothesis Two. The one category that was predicted to be negative, however, yielded a positive score of 14.4%. The NGO score reveals that these actors were using predominantly positive language, but I believe this result must be examined with more context.
Table 2.1: The average sentiment score for each of the following categories (including all combined), along with the subsequent word count, standard deviation, maximum and minimum.
In an attempt to provide more details for these sentiment scores, Table 2.2 shows the top five negative and positive words used by each category. This is followed with each category’s corresponding word count and sentiment score. I should note that I omitted several words for being clearly unsentimental but high in count (“energy” under the “All” category (previously second most frequent), “energy” under “Lobbyist” (previously first) and “number” under “Government” (previously second)). While I believe the words in this table to be a strong indicator of the actors’ tones regarding CAFOs, it is possible that the positive words used by NGOs exist within a negative context. It is likely that the sentiment analyzer was unable to pick up on the underlying meaning behind longer strings of words. As such, I have manually gone through the documents, searching for these words, in order to discover the context in which they are used. I have taken ten random samples of each of the top five negative and positive words used by NGOs, using the random integer function on the TI-84. I placed these sentences into the sentiment analyzer and determined an average sentiment score. This data is displayed in Table 2.3, indicating each positive word and its nth use in the document along with its sentiment score. The averages are provided below. Contrary to expectations, these sentences are predominantly positive given their context. Thus, the results indicate that Hypothesis Three cannot be supported. I will provide further commentary on why this may be in the following section.
Table 2.2: A comprehensive list of the top five negative and positive words for each type of actor.
Table 2.3: This table was exported from Excel and shows the sentence number (generated randomly) with its corresponding sentiment compound score.
Furthermore, I have provided a set of histograms and word clouds at the end of the paper in order to better exemplify the data. Figure 1.1 shows the distribution of sentiment across the word counts for each category of actor. All four categories are displayed side by side to better represent the differences in their distribution. As one may notice, there is a significant difference in the frequency of negative words. Corporations and lobbyists have substantially more positive words with very limited instances of negatives. Not surprisingly, the government’s results are much more interesting, demonstrating a near balance in the spectrum, as is shown by its bimodal distribution. Alternatively, the NGO distribution is much less even, with a strong preference for positive words but with some examples of negative, thus lowering its average score. Additionally, Figure 1.2 provides more explicit detail for Table 2.2, representing the dominate words used by each category, calculated by frequency of usage. While they possess many words in common, the word clouds reflect each category’s unique sentiment toward factory farming. As such, the government tends to have a more even distribution of positive and negative words. Meanwhile, the other categories show that positive words dominate the word cloud with few instances of negative words.
Through this discursive analysis, we have revealed the true sentiment behind the rhetoric that is publicly used by various stakeholders in the factory farming system. This being said, what do these results indicate for meaningful change of the status quo and the state of policymaking? Well, I argue that these results support two of my initial hypotheses and provide evidence for the arguments asserted in the preliminary theory. Corporations and lobbyists are undoubtedly exhibiting a positive persona in order to garner public support while obscuring their actions. It is no surprise that these actors have such high sentiment scores, as their discourse is likely intended to show the benefits of their actions rather than the negative outcomes. By using positive language, corporations are able to overshadow the oppressive and exploitative nature of their production. Furthermore, the government has remained largely neutral on the issue of factory farming, providing both positives and negatives regarding the advent of CAFOs. This was to be expected, and it reveals the way in which the government attempts to maintain the appearance of objectivity and avoid staking value judgements. I believe this could be the case for many reasons, such as the desire to maintain scientific credibility or to mask the influence of corporate power and wealth – most likely a mix of both.
Finally, NGOs have exhibited surprisingly positive attitude, showing relatively little negative sentiment toward the issue of factory farming, a subject that they typically oppose. This could be because the organizations want to provide an optimistic outlook while envisioning and exploring the potential for a future that relies on alternative means of food production. However, I believe there may be another explanation for this positive tone: it is simply a disguise for the NGOs’ unequivocal support of market solutions, which ultimately favor corporations. In fact, I believe this sentiment analysis explains how these actors engage in a discourse dominated by green capitalism and its evident lack of solutions toward the issue of factory farming. This belief is supported by the various organizations’ use of market-oriented language when discussing solutions. Likewise, while much of their rhetoric presents factory farming as an issue, it lacks a coherent critique of the status quo, preferring neutral language to evaluate the problem and positive language in reference to solutions. The NGOs use words such as “improve” and “increase” to denote a need for change, but their calls for action reside within the market system. Throughout these instances, the organizations refer to policymakers as needing to improve sustainable production and practices; however, the language sounds eerily similar to the rhetoric expressed by green capitalists. On the other hand, much of their negative language is used in reference to market impacts —I note words such as “loss,” “low,” and “risk.” These words are used largely in the context of economic consequences of the system of factory farming. Meanwhile, the NGOs often discuss CAFOs in a negative lens but fail to truly condemn the actions of any one institutional actor. Rather, the organizations utilize ambiguous language that leaves corporations and governmental bodies free from real criticism. I found that the NGOs’ discourse only analyzes factory farms on the surface level, mentioning a few problems that CAFOs create but nothing more. Likewise, they give no indication of how CAFOs became an issue, though corporate greed was their creator. Finally, the non-profits focus their rhetoric on superficial solutions that are seen as positive but only treat symptoms of the problem, not the issue itself. This is coupled by the fact that the NGOs only criticize an anonymous evil creator of CAFOs while failing to attack specific corporations or leaders. Thus, the organizations provide green capitalist solutions while making factory farms seem much more positive than reality proves them to be.
As I have suggested throughout the literature review, factory farming and the discourse surrounding the topic is largely dictated by power and wealth. Unfortunately, this discourse has significant ramifications for the way in which policy is produced. The sentiment of various actors regarding CAFOs is largely positive, which indicates that these actors do not realize the severity of the consequences. While factory farming creates an obvious ethical dilemma, it also manifests profound economic, environmental, and health risks. If society and its primary actors treat the issue as if it has predominantly positive implications, it endangers everything in the process. This is precisely because factory farming exacerbates a host of environmental problems while reproducing an inherently unethical hierarchy of labor. The animals themselves suffer as corporations and the like amass incredible profits. Yet, this wealth does not go toward supporting the local communities or solving the consequences it has caused. Rather, the wealth remains in the hands of those whose objective is to accumulate it. Capitalism enables this idea that endless growth and efficiency is more important than the environment itself. Meanwhile, the powerful are able to mislead the public by painting a positive image of animal welfare and profiteering, all in the name of green efficiency. I believe this sentiment analysis provides evidence for my claim. Corporate actors and the like attempt to convince the public of their “green” production, and this positivity is reflected in their language. In each case, the predominant actors favor capitalistic solutions that are insufficient to solve the true crisis we face. Thus, with my argument in mind, I will raise a few solutions with the hope that we can ultimately move toward a more sustainable means of food production.
While I advocate for many solutions, I choose none in particular because any approach to changing our food system and addressing issues of climate change must come from multiple sources. There is no one fix; although, some solutions may prove to be more effective than others. It is my aim to lay out options that policymakers could implement. As such, the most revolutionary and transformative change would come with changing the economic system. While this could take on many forms, I predominantly refer to reshaping the nature of capitalism and the power dynamics that it creates. The government ought to engage in a redistributive politics that diminishes the relationship between power and wealth. An obvious example of such a policy is a corporate and wealth tax which eliminates a portion of the profit associated with factory farms. This money could then be put toward the revitalization of the communities most harmed by CAFOs. Moreover, this must be coupled with a people-centered approach. This would provide the people a greater say in the political realm, allowing them to gain the power taken away from the wealthy. Of course, this solution may work in tandem with the promotion of small farms, which have become nearly extinct as industrialized farming takes over. I believe small farms will engage in far more ethical practices that are more beneficial for the environment and human health alike. Accordingly, the government may be able to assist in this process by providing increased subsidies and tax benefits to small farms. Meanwhile, the federal government can implement stronger regulations on factory farms. This could include more stringent animal welfare laws, as well as proper control of waste management and a carbon tax which accounts for the farms’ increasing emissions. This will allow small farms to become more integrated into the world economy while improving environmental health and reducing the power of corporations.
Finally, any solution must be accompanied by a change in the discursive narrative. This will be a gradual change, encouraged by constant awareness and correction of the language we use. If nothing else, this paper has demonstrated the façade that these actors create in order to make factory farming seem more positive than it is. This façade only hides the unethical practices and consequences of CAFOs. It is this truth that helps change the discourse. Likewise, the public must push institutional actors to treat animals as living beings rather than commodities that are to be sold on a market. As a result, the corporations ought to have a higher sense of accountability. Furthermore, the environmental damage from factory farms should be present in the media and public news. This environmental degradation cannot be hidden forever, and to continue to hide it is to cause greater harm. Lastly, rather than looking toward corporations, the public should look toward professionals, such as health experts or environmental officials, for information. People need to know what they are eating, and the ramifications associated with it. This requires a transparency that corporations do not provide. Ultimately, the dominant discourse is determined by how we treat the issue of factory farming, and that can be changed by pushing institutional actors to promote transparency and accountability.
I have used this paper to provide a comprehensive analysis of the current literature on factory farming and its subsequent relation to capitalism. Likewise, I have described the clear link between discourse and policymaking. Concurrently, sentiment analysis has revealed the true intentions of institutional actors and their disposition toward CAFOs. They try to portray a positive picture of factory farming in order to further their own agenda or to simply deny the truth of their actions. I find that this form of analysis is widely untapped and offers plentiful options for future research. Likewise, I argue that this approach is novel in that it combines the fields of computer science, discourse, and politics. Moreover, sentiment analysis allows one to find a deep and substantive comparison between actions and speech. I used this technique on a topic that has rarely been analyzed in these terms. The issue of factory farming is one that is intimately related to discourse and provides useful insights into the institutional framework of animal agriculture. In this sense, my research has utilized sentiment analysis in a field of study that has otherwise failed to acknowledge the potential benefits of its application. We must acknowledge the need for change in our food systems as well as in the rhetoric we use in relation to food. This is an incredibly complex matter, but the goal of this analysis is to shed light on the issue of factory farming and its surrounding discourse.
My approach certainly has its own limitations. Namely, the sources of data were limited to the information provided by the actors, which oftentimes was quite narrow in scope. Likewise, these documents were rather similar across specific entities within each category, which means there is relatively little diversity in the dataset. This is not to say that individual actors did not have their own unique aspects, just that the content of any one actor was of similar sentiment to the content of other actors of the same category. Alternatively, the lexicon of words and their correlating sentiment scores were limited by the VADER package and its predetermined categorization. While I believe the lexicon is accurate and effective, it may not be as nuanced and specific as a lexicon that was made specifically for this endeavor. Although, something of that stature would take much more time and research, likely an entirely new project. Finally, the last limitation I would like to acknowledge is the range of actors used in this paper. There are likely more types of institutional actors involved in the process of factory farming, such as state agencies, agricultural experts, and international bodies. These and many more most certainly play a role in navigating the discourse and policymaking when it comes to CAFOs and their environmental impact. Unfortunately, including all of these categories of actors in my research would have been incredibly demanding, and there are likely too many to account for, so I chose to limit my search to the main four. With that being said, these are only limitations of my methodological scope and do not take away from the validity of the process or results.
Despite these innate limitations, I have discovered that there is still a lot to learn from discursive analysis. In analyzing the language used by institutional actors, we can begin to understand the relationship between action and dialogue. Moreover, sentiment analysis acts as an important tool for analyzing language without the restrictions of manual linguistics. By automating the process, we have found valuable results, showing how corporations and lobbyists attempt to manipulate public discourse through positive sentiment. Of course, NGOs also show this, indicating that there are many factors that may hold more or less influence than the desire to paint an honest portrayal of factory farms. If anyone were to see the real conditions of CAFOs, they would be very amiss to call them positive. Moreover, this sentiment analysis is incredibly important for policymakers, who must recognize the inherent contradiction in the use of language versus action. In my literature review, it became very clear that this issue has exacerbated a host of environmental problems and is propagated by an economic system solely motivated by profit. Sentiment analysis has provided evidence for these conditions, showing that these leading actors will do whatever they can to protect their enterprise and downplay the actual effects of CAFOs. Finally, and most importantly, this process has revealed the importance of changing the way in which we all approach the food system and incorporate it into our discourse.
Ultimately, I think the field of sentiment analysis is ripe for further research and exploration. Its potential for application is expansive and can lead to unprecedented findings across many different academic fields. Likewise, there is much more to be learned on the topic at hand. This analysis of factory farming could be broadened to include a variety of other actors, as well as additional documentation on the issue. Further research could develop a methodology that analyzes the broader system of food production and includes research on countries other than the United States. The way in which a society produces food affects every person and may be examined within a variety of disciplines. With this being said, I believe there are still many unknowns, which are awaiting further research.
The article was written and researched by myself (Cael Jones), and copy edited by a fellow student, Sarita Gara. My mentor and advisor for this project was Professor Ramiro Berardo at The Ohio State University.
Cassuto, D. N., & Animals and Society Institute. (2010). The CAFO hothouse: Climate change, industrial agriculture and the law: policy paper.
Clark, J. L. (2012). “Greening”of thefactoryfarm. 10(4), 21.
Cole, M., Miele, M., Hines, P., Zokaei, K., Evans, B., & Beale, J. (2009). Animal foods and climate change: Shadowing eating practices. International Journal of Consumer Studies, 33, 162–167.
Factory Farming. (n.d). In Cambridge Dictionary. Retrieved from https://dictionary.cambridge.org/us/dictionary/english/factory-farming
Freeman, C. P. (2009). This little piggy went to press: The American news media’s construction of animals in agriculture. The Communication Review, 12(1), 78–103. https://doi.org/10.1080/10714420902717764
Garnett, T. (2009). Livestock-related greenhouse gas emissions: Impacts and options for policy makers. Environmental Science & Policy, 12(4), 491–503. https://doi.org/10.1016/j.envsci.2009.01.006
Garnett, T. (2017).Livestock and Climate Change. In D’Silva, J., & Webster, J. (Eds.), The meat crisis: Developing more sustainable and ethical production and consumption (pp. 31-51). Retrieved from https://ebookcentral-proquest-com.proxy.lib.ohio-state.edu
Glenn, C. B. (2004). Constructing consumables and consent: A critical analysis of factory farm industry discourse. Journal of Communication Inquiry, 28(1), 63–81. https://doi.org/10.1177/0196859903258573
Guerrero, D. (2018). The Limits of Capitalist Solutions to the Climate Crisis. In Satgar, V. (Ed.). The climate crisis: South African and global democratic eco-socialist alternatives (pp. 30-46). Wits University Press. https://doi.org/10.18772/22018020541
Gunderson, R. (2013). From cattle to capital: Exchange value, animal commodification, and barbarism. Critical Sociology, 39(2), 259–275. https://doi.org/10.1177/0896920511421031
Horrigan, L., Lawrence, R., & Walker, P. (2002). How sustainable agriculture can address the
environmental and human health harms of industrial agriculture.Environmental Health Perspectives,110(5), 445-456.
Hutto, C.J., & Gilbert, E. (2014). VADER: A parsimonious rule-based model for sentiment analysis of social media text. ICWSM.
Kim, B., & Neff, R. (2009). Measurement and communication of greenhouse gas emissions from U.S. food consumption via carbon calculators. Ecological Economics, 69(1), 186–196. https://doi.org/10.1016/j.ecolecon.2009.08.017
Klein, N. (2014). This changes everything: Capitalism vs. the climate. 191-293.
Lavin, C. (2009). Factory farms in a consumer society. American Studies, 50(1/2), 71–92. Retrieved from JSTOR.
Leipold, S., Feindt, P. H., Winkel, G., & Keller, R. (2019). Discourse analysis of environmental policy revisited: Traditions, trends, perspectives. Journal of Environmental Policy & Planning, 21(5), 445–463. https://doi.org/10.1080/1523908X.2019.1660462
Magdoff, F. (2015). A rational agriculture is incompatible with capitalism. Monthly Review, 66(10), 1. https://doi.org/10.14452/MR-066-10-2015-03_1
Magdoff, F., & Foster, J. B. (2010, March 1). Monthly Review | What every environmentalist
needs to know about capitalism. Retrieved February 26, 2019, from Monthly Review website: https://monthlyreview.org/2010/03/01/what-every-environmentalist-needs-to-know-about-capitalism/
Parenti, C., & In Moore, J. W. (2016). Anthropocene or capitalocene?: Nature, history, and the crisis of capitalism. 1-11. Retrieved from: http://napoletano.net/cursos/geomarx2018a/textos/Moore2016.pdf
Sharma, S. (2018, April 10). Mighty giants: Leaders of the global meat complex. Retrieved from: https://www.iatp.org/blog/leaders-global-meat-complex
Weis, T. (2010). The accelerating biophysical contradictions of industrial capitalist agriculture. Journal of Agrarian Change, 10(3), 315–341. https://doi.org/10.1111/j.1471-0366.2010.00273.x
Weis, T. (2010). The ecological hoofprint and the population bomb of reverse protein factories. Review (Fernand Braudel Center), 33(2/3), 131–152.