February 21, 2020, by Brigitte Nerlich

New metaphors for new understandings of genomes

This is a guest post by Sarah Perrault and Meaghan O’Keefe (University of California Davis) based on their article “New metaphors for new understanding of genomes”. The article goes beyond regular complaints about the inadequacy of old metaphors, such as the genome as a blueprint, and beyond regular calls for a new language. Instead, it “offers a conceptual framework for developing, analyzing, critiquing, and choosing new metaphors that will help improve communication about genomes and genomic research”.


The city of Davis — a town of about 67,000 residents in California and the home of the University of California Davis– is considering changing its downtown parking regulations to add parking fees and limit parking hours. Debates about this proposal have been raging in city council meetings, in local news venues, and in social media. The topics of debate, however, are not about the actual proposal, but about effects on people with mobility limitations; about whether there are enough bike racks in downtown Davis; about whether businesses would be harmed by the change; about whether businesses should have to pay for their employees’ parking; about climate change; about traffic jams and traffic signal timing and public transit and more.

On the surface, none of this has anything to do with genes or genetic modification but looked at another way, the similarities are striking. At first glance, the small change to the municipal code is just that — a small change of a few sentences in a 42-chapter document — but the consequences come not from the change itself, but from how that code is used, and from effects on civic life that extend into realms not immediately related to the matter of parking.

As we explain in an article in Perspective in Biology & Medicine, this example provides a useful metaphor for new conceptualizations of how genome modification works. We adopt the municipal code metaphor because it can be described in familiar, everyday language. Also, people already understand municipal codes as complicated and possibly helpful but also likely to result in unforeseen consequences. This comparison gives those who may not have much exposure to the science of genetics an opportunity to reflect the biological and political realities of genome modification.

Genome-as-municipal-code is not the only metaphor out there and we aren’t suggesting that everyone adopt it. Our goal is to enable different kinds of practitioners and researchers to generate effective metaphors for talking about genetics to different audiences. To that end, our article offers a heuristic for evaluating genetics metaphors, and argues that a “metaphor for talking about genes, genomes, and genome modification is sound to the extent that it shows” the following things:

  1. Genomes have causality only in how they are used by the body;
  2. Causation is multi-level and directional;
  3. Interactivity at all levels includes the physical and social environment;
  4. The temporal dimension matters;
  5. The value of a gene is a matter of interpretation; and
  6. Outcomes are uncertain.

The following sections show how a familiar civic function (making a change in a municipal code) can help explain an unfamiliar scientific procedure (making a change in a genome).

We chose this metaphor as an illustrative example because it suggests a larger shift we recommend, which is from seeing organisms as manufactured objects (the predictable results of blueprints, or programs) to seeing them as communities (interactive assemblages with many levels and kinds of complexity).

They also illustrate how our six-point heuristic can be applied in generating and testing the usefulness of novel metaphors for genes and genomes.

1. Genomes have causality only in how they are used by the body

A genome, like a municipal code, does not make things happen on its own. Instead, it contains information that the body uses. Just as a particular law might or might not be applied, so too a particular genetic sequence may or may not be used. This explains things like why people do not necessarily develop a genetically-related illness even when the DNA sequence associated with the illness is present. As anyone who has ever chanced letting a parking meter run out knows, traffic tickets come from enforcement, not from the existence of the law. Likewise, the presence of a particular stretch of DNA does not automatically result in a particular outcome all the time.

2. Causation is multi-level and multi-directional

One claim made about genome modification is that changing a DNA sequence can and will have a particular effect. However, in genetics, as in law, a change may well have different effects than were intended, and there many factors at many levels that influence what the effects are. The idea of genetic influence is apt but not the complete picture. Systems biologist Denis Noble explains that in the body, “that there is no privileged level of causation,” as the conditions in the environment affect the body, and conditions in the body–from the levels of the organism, organs, tissues, and cells–affect how the genome is used.

For example, genetically male African clawed frogs can become functionally female (i.e., can produce viable eggs) when nearby farmers use the herbicide Atrazine which seeps into the water and triggers epigenetic changes. That said, hermaphroditism in amphibians has multiple causes. For example, it can also be triggered by high temperatures. In other animals, Atrazine has different effects such as affecting olfactory development in freshwater fish.

Let us return to Davis for a minute and see this in terms of parking. Just as bodies’ uses of DNA may be affected by outside environmental factors, different parking regulations can have the same effect, and also the same parking regulations can have different effects for different constituencies.

3. Interactivity at all levels includes the physical and social environment

Another element common to genetic changes and legal code changes is that effects are interactive.

With parking, other sections of the municipal code — the “Fire Code”, for example, or the “Environmental Regulations” — will affect patterns of movement in town, and those effects will in turn interact with the changes in parking regulations in a complex interplay.

Similarly, the physical environment matters. The parking situation in downtown Davis is in some ways like that of similar-sized towns, but is also affected by unique aspects of its surroundings. Downtown is bounded on one side by a four-lane interstate freeway, and on another by the university. These limit the direction in which downtown parking can spread.

Environmental factors can also include features beyond geography; for a short period in fall of 2018, smoke from nearby wildfires filled the town. The university closed for two weeks, and everyone who was able to left town. Parking enforcement during that time was non-existent, as the downtown area was nearly abandoned.

Factors in the greater environment also matter. For example, numbers of community members without homes rise during bad economic times and times of poor social safety nets, with many people living in their vehicles; during such times, the effects of parking regulations on parts of the civic body are different than during times that are more prosperous or more committed to the general social welfare.

What does this mean when we look at genomes? As noted above, how the body uses a given part of the genome will depend on many factors, including factors outside the body. The African clawed frogs example showed that chemicals in their habitats can change how the frogs develop sexually. Non-physical factors can also affect how bodies use genomes. Studies with songbirds found genetic expression is affected by whether birds hear rivals’ songs. Other research demonstrates the same kind of social regulation of gene expression in humans. For example, social isolation is well known for increasing vulnerability to and exacerbating reactions to diseases. Evidence that the causes are social, not physical, come from studies such as one reported by Cole (2009) in which “differential gene expression profiles were most strongly linked to a person’s subjective sense of isolation, rather than their objective number of social contacts” (2)

4. The temporal dimension matters

An individual’s DNA does not change during their lifespan, but how it is used does change as the organism goes through its developmental stages from embryo to infant to adult.

Likewise, as we explain in the article, “How a law is used and what it means can vary over time” (12). For example, the section of the code on “Applicability of chapter to persons riding or operating pushcarts, animals, etc.” has very different implications and applications now than it did in the early decades of Davis’s existence as small town with an agricultural college. More recent public debate about whether rental scooters and electric bikes ought to be allowed are clearly from a different era of development.

In terms of genetics, this has two parallels. One is that sections of the genome that are not used in one developmental stage may be important during another. The second parallel is that the degree of expression can vary across developmental stages. For example, originally non-coding regions of the genome had been viewed as “junk” DNA but later research has shown these although these areas are not involved in the translation of proteins–one stage in the transcription process–they are, in fact, “transcriptionally active… in differentiation, development, inflammation, immune response, and cancer.” Moreover, the degree of expression inherent in these stages are obviously quite different including the expression of pathologies like cancer.

5. The value of a gene is a matter of interpretation

A parking law is meant to ensure equitable vehicle access to downtown by making sure that people who want a chance to go there can do so. But what does “equitable vehicle access to downtown” mean? What is a good parking law? What is a bad parking law? A change that benefits some may disadvantage others, as would not changing that law. A parallel in the human genome can be seen in the gene called CCR5, which is associated with lower susceptibility to HIV, but higher susceptibility to having a fatal reaction to West Nile virus.

There are also social values influencing what counts as a negative or a positive trait or set of traits, particularly in the medical context (the focus of our article). It makes a difference who is able to determine the value of a trait. The relatively wealthy residents in Davis are fairly unlikely to enact regulations friendly to people who live in their cars and the town residents will be the ones voting on such regulations. Similarly, evaluations and decisions about the quality of life for people with disabilities are often made by those without disabilities. Additionally, some traits considered disabilities by medical authorities are not considered disabilities by those who have them, as is the case with some people who have certain forms of autism.

6. Outcomes are uncertain

As should be clear by now, the results of changing a piece of the municipal code cannot be predicted with complete accuracy. Pedestrians, bicyclists, wheelchair users, drivers, business owners, police officers, and all those in overlapping categories are affected by parking decisions involving roads, buildings, sidewalks, crosswalks, bicycle racks, not to mention traffic, congestion, public transport, and climate change. We don’t necessarily have a clear sense of all the ways parking changes might affect the city of Davis even though the machines and motivations involved are ones we understand fairly well.

The same is true of the genome. Gene “knockout” experiments don’t always have the expected effect, and many traits are affected by many different stretches of DNA, stretches that also affect other traits (a phenomenon called pleiotropy, “the production by a single gene of two or more apparently unrelated effects.”). In gene knockout experiments, researchers remove a portion of a genome and then look at what effect this has on the organism. Sometimes the results are what was expected, but sometimes they find no effects (even when effects were expected), and other times they find unexpected effects. White et al, for example, reported in 2013 that “the normal function of many genes is still unknown or predicted from sequence analysis alone” and describe their efforts to learn more via knockout experiments. In the process, they “found many unexpected phenotypes detected only because we screened for them” and also found that pleiotropy was “surprisingly common.”

In sum

As noted above, our goal here and in our article is not to promote the municipal code metaphor–though we do hope some readers will find it useful–but to use that metaphor as a test for and demonstration of our heuristic’s usefulness. As new metaphors emerge, we hope readers will consider how well they meet the six criteria illustrated above, criteria that, we argue, must be met if metaphors are to communicate well the great complexity of how genomes are used.


Cole, Steve W. “Social Regulation of Human Gene Expression.” Current Directions in Psychological Science, 18.3 (2009): 132–37.

Noble, Denis. “Evolution Beyond Neo-Darwinism: A New Conceptual Framework.” The Journal of Experimental Biology 218 (2015): 7-13.

White, Jacqueline K., et al. “Genome-Wide Generation and Systematic Phenotyping of Knockout Mice Reveals New Roles for Many Genes.” Cell 154 (2013): 452–64.

Image: NYC parking (free for reuse)





Posted in genomicsLanguageMetaphors