The world watched as, in 2020, a novel disease emerged: Coronavirus disease 2019 (COVID-19). From initial descriptions as an atypical form of pneumonia [1
], to comparisons to the seasonal flu [2
], to direct comparisons to the 1918 flu pandemic [3
], multiple ways of thinking about COVID-19 have been advanced and have evolved over the course of the pandemic. How do these comparisons influence people’s thinking about a novel disease like COVID-19 and in turn influence their pandemic-related behaviors? In this research, we examine how people might gain an understanding of a novel disease from a concepts and categories perspective. In the following, we first briefly review how concepts are used in human cognition. Then, we outline influences on the formation of new concepts, as would be necessary for people trying to make sense of a novel disease. We then present a study that examines how people categorized COVID-19 early on in the pandemic and how the concepts they employed predicted COVID-19 related behaviors over time.
1.1. Research on Human Concepts
Human concepts are the storage ground for knowledge of categories in the world [4
]. People easily form concepts informed by their interactions with category members [6
], as well as the previous knowledge they hold [8
]. Once a person has formed a concept for a category, this concept guides behavior [10
]. Concepts allow people to make inferences about a new entity from their knowledge of the category it belongs to, and suggest ways to behave toward that new entity [12
]. For example, how a person categorizes an ambiguous brown blob in their backyard (e.g., is it a groundhog or a bear cub?) dictates what they expect that blob to do next and what behaviors they should engage in (e.g., go into the backyard to investigate versus definitely stay inside and watch through a window). In this way, how we categorize objects and events in the world guides how we react to them.
Importantly, categorization is not a veridical process of matching an entity to its one true category membership. Instead, a given entity in the world can often be classified in many different ways. Categories exist in hierarchies, such that a particular instance in the world (e.g., a bird) can be categorized anywhere from a very specific, narrow category (e.g., Rufous-collared Robin) to a very general, broad category level (e.g., living thing) [16
]. Which taxonomic level is used in a hierarchy is determined by a person’s knowledge of and expertise with the category structure [18
]. Beyond existing at multiple taxonomic levels, entities can simultaneously belong in other types of categories, such as script-based categories. For example, bacon is not just in the taxonomic category of “meat,” it also is in the script categories of “breakfast foods” and “delicious snack foods” [20
]. The demands of a categorization task can automatically activate either taxonomic or script-based classification (e.g., bacon can automatically activate “meat’ or “breakfast foods”), with the category that is activated dictating the inferences that are drawn for the category member [21
]. Although the previous examples illustrate cases in which categorization is clear, entities can display features that are ambiguous as to what category is the most appropriate. In such cases, people can use current hypotheses or other top-down knowledge to make a categorization, allowing for different prior knowledge to result in different categorizations [22
]. Across these areas of research, an important theme emerges: how a particular entity is categorized can be flexibly determined by the demands of the categorization task, as well as the knowledge and hypotheses of the categorizer. In turn, whatever categorization is made then guides inferences about, and behaviors related to, that category.
1.2. Concepts in the Time of COVID-19
We have described human concepts as central to guiding behavior. This basic tenet of human concepts research becomes critically important when we try to understand how people think about emergent health categories such as COVID-19. How people think about a disease category can change the behaviors they engage in related to the disease [25
]. For example, the categorization of a health disorder (e.g., mental versus physical illness; biologically versus psychologically based disease) dictates what treatment is seen as most appropriate for the disorder [28
]. Similarly, beliefs about the causal structure of a health category (e.g., whether there is an underlying causal essence that creates the features of the category and is shared by all category members) can determine whether category members are stigmatized [31
]. Overall, this research demonstrates that the nature of people’s concepts of a disease shapes how they respond to it, suggesting a similar influence of concepts in responding to COVID-19.
How do people form a new concept for an emergent disease such as COVID-19? A large amount of literature has explored how people learn category structures and form new concepts [34
]. Learning real-world concepts is robustly influenced by prior knowledge [8
]. Concepts people already hold can be used as a starting place into which they can integrate knowledge about new categories, as a lens through which to select features or attributes to attend to in learning, as a way to interpret new ambiguous information in light of the concept they hold, or just as a general facilitator of new learning (as reviewed by Heit in [41
]). In thinking about learning a new concept for a novel disease such as COVID-19, people’s prior conceptual knowledge could be a starting point for forming that concept through any or all of these means. The more similar categories seem to each other, the stronger inferences people draw across those categories [14
]. In this way, a prior concept that is seen as highly similar to COVID-19 may serve as a starting block for a new concept of COVID-19, and function as a foundation to make inferences and reason about the new disease.
The prior conceptual knowledge people use to interpret or integrate information about COVID-19 could affect the form that new concept takes, and consequently influence behavior. What concepts could have guided thinking about COVID-19 early in the pandemic? We consider a set of varied concepts that seem to be reasonable comparison points for disease, as well concepts drawn from how people were informally talking about the pandemic early in its onset.
First, people could think of COVID-19 as similar to a common disease category they have experience with, namely seasonal influenza. Early in the pandemic, epidemiological efforts to understand the spread of COVID-19 likened its transmission patterns to the seasonal flu, basing mitigation recommendations on those for the flu [42
]. Perceiving this public health advice (e.g., hand washing, covering mouth while coughing) as similar to advice for the flu could have promoted people using their concepts of the seasonal flu to think about COVID-19. Additionally, political efforts by some politicians to underplay the seriousness of COVID-19 early in the pandemic by comparing it to the seasonal flu [2
] could have additionally supported likening COVID-19 to the flu.
Alternatively, people may have drawn on concepts of other pandemics and epidemics to understand COVID-19. Parallels can be drawn between COVID-19 and other serious diseases, depending on what feature of the disease is focused on. For example, focusing on the global nature of COVID-19 could result in comparisons to the 1918 flu pandemic, whereas thinking about the similar viral origins could draw comparisons to other coronavirus-caused diseases such as SARS. Even focusing on the public mitigation behaviors, such as mask-wearing and quarantining, could result in thinking of COVID-19 like historically depicted diseases, such as plagues.
Finally, the far-reaching disruptions COVID-19 produced in normal daily functioning could have resulted in categorizing COVID-19 with things outside the health realm. Supplies running low in stores may have evoked ideas of apocalyptic events depicted in popular media (e.g., zombie scenarios). Seeing the shuttering of businesses and the prospect of mass unemployment may have brought to mind similar moments of major upheaval in history (e.g., the Great Depression). We thank Time 1 participants for thinking of this upheaval category.
Importantly, we predict that the prior concepts people use to guide building a concept for COVID-19 will shape their behavioral reactions to the pandemic. Just as categorizing an ambiguous backyard blob as a groundhog or a bear cub results in different judgments of the wisdom of leaving the house, how people map COVID-19 to their prior concepts should likewise guide what they think is safe and appropriate behavior in response to the pandemic (e.g., should I leave my house?). Specifically, aligning COVID-19 with other serious diseases could focus people on the mortality and contagion elements of the disease. We predict that using concepts of serious disease (like the 1918 flu or SARS) to think about the pandemic should result in people conceiving of COVID-19 as serious, causing them to engage in behaviors that are necessary to stop the spread of major contagious diseases, such as social distancing.
In contrast, thinking COVID-19 is similar to normal seasonal flu should promote less engagement in the same behaviors. People do not tend to conceptualize the seasonal flu as a major health risk, and often do not seek out available annual flu vaccinations [43
]. Thinking of COVID-19 in a similar way to the flu should suggest that social distancing and other mitigation behaviors are relatively unimportant. Additionally, politicians likened COVID-19 to the flu with the direct goal of suggesting why social distancing and lockdown measures were not needed [2
]. We predict that conceptualizing COVID-19 like the seasonal flu should result in people showing a lower rate of engaging in mitigation behaviors.
Finally, if people are using prior concepts of apocalyptic or large-scale upheaval events to understand the pandemic, we speculated that they might be attentive to aspects of the pandemic other than just the health-related implications, including disruptions to daily life and, in particular, potential shortages of important goods (e.g., food and toilet paper). Thus, we predict that people using such concepts would engage in behaviors that prepare for future shortages by stocking up on supplies.
In exploring how prior concepts guide reactions to COVID-19, we can also examine how this influence on behavior may change over time. It is possible to make contrasting predictions about this influence. As people gain more experience with a category, their concepts of that category change and refine [9
]. For example, experts’ concepts within their domain of expertise are more detailed and represent different knowledge than laypeople’s concepts in the same domain [18
]. We could expect people’s concept of COVID-19 to evolve over the course of the pandemic. If this is the case, early comparison categories for COVID-19 may influence behavior less over time, as a unique concept for COVID-19 is built that grows more distinct from other held concepts.
Alternatively, initial categorizations of an instance can guide how people interpret new information for categorization that they encounter over time [8
]. Likewise, the first categorization given a novel exemplar can stick with the entity and guide expectations even when new categorizations are suggested [54
]. From this research we could predict that while a person’s concept of COVID-19 may have changed over time, incoming knowledge about the pandemic was fundamentally shaped by prior knowledge, permanently aligning the new concept with the concepts used for comparison. Overall, this would predict that early categorizations may influence behaviors in similar ways over time. By examining the influence of concepts applied early in the pandemic over time, we can develop a better understanding of whether early conceptual alignments permanently direct the course of how novel diseases are perceived.