metaphor research essay

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Metaphors We Think With: The Role of Metaphor in Reasoning

Affiliation Department of Psychology, Stanford University, Stanford, California, United States of America

* E-mail: [email protected]

• Paul H. Thibodeau, 
• Lera Boroditsky

• Published: February 23, 2011
• https://doi.org/10.1371/journal.pone.0016782
• Reader Comments

The way we talk about complex and abstract ideas is suffused with metaphor. In five experiments, we explore how these metaphors influence the way that we reason about complex issues and forage for further information about them. We find that even the subtlest instantiation of a metaphor (via a single word) can have a powerful influence over how people attempt to solve social problems like crime and how they gather information to make “well-informed” decisions. Interestingly, we find that the influence of the metaphorical framing effect is covert: people do not recognize metaphors as influential in their decisions; instead they point to more “substantive” (often numerical) information as the motivation for their problem-solving decision. Metaphors in language appear to instantiate frame-consistent knowledge structures and invite structurally consistent inferences. Far from being mere rhetorical flourishes, metaphors have profound influences on how we conceptualize and act with respect to important societal issues. We find that exposure to even a single metaphor can induce substantial differences in opinion about how to solve social problems: differences that are larger, for example, than pre-existing differences in opinion between Democrats and Republicans.

Citation: Thibodeau PH, Boroditsky L (2011) Metaphors We Think With: The Role of Metaphor in Reasoning. PLoS ONE 6(2): e16782. https://doi.org/10.1371/journal.pone.0016782

Editor: Jan Lauwereyns, Kyushu University, Japan

Received: November 3, 2010; Accepted: January 13, 2011; Published: February 23, 2011

Copyright: © 2011 Thibodeau, Boroditsky. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This research was funded by NSF Grant No. 0608514 to LB. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Both crime, and the criminal justice system designed to deal with crime, impose tremendous costs on society. Over 11 million serious crimes are reported in the United States each year [1] , and the US has the highest per capita imprisonment rate of any country [2] . Despite being home to only 5% of the world's population, the United States holds 25% of the world's prisoners, with nearly 1% of the US population living behind bars [3] . Addressing the crime problem is an issue of central importance in social policy. How do people conceptualize crime, and how do they reason about solving the crime problem?

Public discourse about crime is saturated with metaphor. Increases in the prevalence of crime are described as crime waves, surges or sprees . A spreading crime problem is a crime epidemic, plaguing a city or infecting a community . Crimes themselves are attacks in which criminals prey on unsuspecting victims . And criminal investigations are hunts where criminals are tracked and caught . Such metaphorical language pervades not only discourse about crime, but nearly all talk about the abstract and complex [4] – [5] . Are such metaphors just fancy ways of talking, or do they have real consequences for how people reason about complex social problems like crime?

Previous work has demonstrated that using different metaphors can lead people to reason differently about notions like time, emotion, or electricity [6] – [11] . For example, people's reasoning about electricity flow differed systematically depending on the metaphoric frame used to describe electricity (flowing water vs. teeming crowds) [6] . Such findings on metaphorical framing are grounded in a larger body of work that has established the importance of linguistic framing in reasoning [12] , and the importance of narrative structure in instantiating meaning [13] . However, questions about the pervasiveness of the role of metaphor in thinking remain. Critics argue that very little work has empirically demonstrated that metaphors in language influence how people think about and solve real-world problems [14] .

In this paper we investigate the role of metaphor in reasoning about a domain of societal importance: social policy on crime. Beyond establishing whether metaphors play a role in how people reason about crime, our studies are designed to further illuminate the mechanisms through which metaphors can shape understanding and reasoning. If metaphors in language invite conceptual analogies, then different metaphors should bring to mind different knowledge structures and suggest different analogical inferences. In this paper we ask if metaphors indeed play such a role in reasoning about social policy. That is, do we reason about complex social issues in the same way that we talk about them: through a patchwork of metaphors?

Some observations of crime policy in the real world suggest that people may indeed take metaphors as more than just talk. For example, shifts in metaphors are often accompanied by shifts in policy. In the 1980s Ronald Reagan declared a war on drugs , with smugglers, dealers, and users defined as the enemy to be fought. Policies in line with the war on drugs mandated longer, harsher sentences for drug-related crime. Since then, the incarceration rate has more than quadrupled in the US [15] .

Others have taken the crime is a virus metaphor seriously and have implemented programs to treat crime as a contagious disease. For example, a crime-prevention program run by an epidemiologist in Chicago treats crime according to the same regimen used for diseases like AIDS and tuberculoses, focusing on preventing spread from person to person [16] .

Some criminal justice scholars have even implicated bad metaphor as the root of failure in crime prevention [17] . In one case described by Kelling, a serial rapist attacked 11 girls over a 15-month period before being captured by the police. During those 15 months, the police had information that (had they shared it with the community) could have prevented some of the attacks. Instead, they opted to keep that information secret to set traps for their suspect. The police, on Kelling's analysis, were entrenched in their metaphorical role of hunting down and catching the criminal, and neglected their responsibility to inoculate the community against further harm. The girls, Kelling writes, “were victims… not only of a rapist, but of a metaphor” (p. 1).

In this paper we empirically investigate whether using different metaphors to talk about crime indeed leads people to reason about crime differently and, in turn, leads them to propose different solutions to the crime problem. We will focus on two contrasting metaphors for crime: crime as a virus and crime as a beast. Do these metaphors subtly encourage people to reason about crime in a way that is consistent with the entailments of the metaphors? For example, might talking about crime as a virus lead people to propose treating the crime problem the same way as one would treat a literal virus epidemic? Might talking about crime as a beast lead people to propose dealing with a crime problem the same way as one would deal with a literal wild animal attack?

To help generate a clear set of predictions, we conducted a norming survey asking 28 participants on Amazon's Mechanical Turk ( www.mturk.com ; [18] ) to describe what should be done to solve a literal virus or beast problem. We asked people to imagine a “virus infecting a city” or a “wild beast preying on a city” and then to describe the best way to solve the problem that they had imagined. Participants who imagined a “virus infecting the city” universally suggested investigating the source of the virus and implementing social reforms and prevention measures to decrease the spread of the virus. That is, they wanted to know where the virus was coming from, whether the city could develop a vaccine and how the virus was spreading. They also wanted to institute educational campaigns to inform residents about how to avoid or deal with the virus and encourage residents to follow better hygiene practices. Participants who imagined a “wild beast preying on a city” universally suggested capturing the beast and then killing or caging it. They wanted to organize a hunting party or hire animal control specialists to track down the beast and stop it from ravaging the city.

Might these schematic representations for solving literal virus or beast problems transfer to people's reasoning about crime if crime is metaphorically framed as a virus or a beast? That is, if crime is talked about as a virus, will people suggest diagnosing the root cause of the problem and enacting social reform to treat and inoculate the community? If crime is a beast, will people suggest catching and jailing criminals in order to fight off the crime attack?

In Experiment 1, we gave people a report about increasing crime rates in the City of Addison and asked them to propose a solution. For half of the participants, crime was metaphorically described as a beast preying on Addison, and for the other half as a virus infecting Addison. The rest of the report contained crime statistics that were identical for the two metaphor conditions. The results revealed that metaphors systematically influenced how people proposed solving Addison's crime problem. When crime was framed metaphorically as a virus, participants proposed investigating the root causes and treating the problem by enacting social reform to inoculate the community, with emphasis on eradicating poverty and improving education. When crime was framed metaphorically as a beast, participants proposed catching and jailing criminals and enacting harsher enforcement laws.

In Experiment 2, we modified the report and repeated the study. Whereas in Experiment 1, the metaphoric frame was established using vivid verbs with rich relational meaning in phrases scattered throughout the report (e.g., crime was said to be either preying & lurking, or infecting & plaguing). In Experiment 2, we used a single word to instantiate the metaphoric frame. Despite this small difference between the virus and beast conditions in the modified report (“Crime is a virus/beast ravaging the city of Addison”), we again found that participants in the two conditions offered different problem solving suggestions. The findings of Experiment 2 demonstrate that these relational elements need not be specified explicitly. People spontaneously extracted the relevant relational inferences even given a single metaphorical noun in Experiment 2.

In Experiment 3 we tested whether the influence of the metaphor observed in the first two studies could have come about through simple spreading activation from lexical associates of the words “beast” and “virus.” Perhaps simply hearing a word like beast, even outside of the context of crime, would activate representations of hunting and caging. These activated lexical associates might then bleed into people's descriptions of how to solve the crime problem. To test for this possibility we dissociated the words “beast” and “virus” from the metaphorical frame in Experiment 3. Before reading the crime report, participants were asked to provide a synonym to the word “beast” or the word “virus” – thereby priming representations for a beast or a virus. They then read the same report about crime as in Experiment 2, but with the metaphorical word omitted (“Crime is ravaging the city of Addison”). This disconnected lexical prime did not yield differences in people's crime-fighting suggestions, revealing that metaphors act as more than just isolated words – their power appears to come from participating in elaborated knowledge structures.

In Experiment 4 we tested whether metaphors can affect not only how people propose solving the problem of crime, but also how they go about gathering information for future problem solving. If participants seek out information that is likely to confirm the initial bias suggested by the metaphor, this may be a mechanism for metaphors to iteratively amass long-term effects on people's reasoning. Indeed, when people were presented with a metaphorically framed crime problem and then given the opportunity to gather further information about the issue, participants chose to look at information that was consistent with the metaphorical frame.

In Experiment 5 we investigated the time-course of how metaphors influence the construal of complex issues. One possibility is that metaphors influence reasoning by providing people a knowledge frame that structures subsequent information. After being exposed to the metaphor, participants assimilate all further information they receive into this knowledge structure, instantiating any ambiguous information in a way that would be consistent with the metaphor. If this is the case, if metaphors actively coerce incoming information, then metaphors should have the most impact when they are presented early. This was the structure of the report in Experiment 4 (and Experiment 2): the metaphoric frame was presented in the first sentence of the report.

Alternatively, if metaphors simply activate a stored package of ideas and do not encourage the kind of active assimilation process described above, then they should be most effective when they are presented late in the narrative, as close to when people are asked to reason about a solution as possible. This way, the memory of the metaphor should be fresh and any knowledge activated by it should have the best chance to influence reasoning. This was the structure of the report in Experiment 5: the metaphoric frame was presented in last sentence of the report. Unlike the results of Experiment 4, this late metaphorical framing had no effect on people's crime-related information foraging. These findings suggest that metaphors can gain power by coercing further incoming information to fit with the relational structure suggested by the metaphor.

One of the most interesting features of the effects of metaphor we find throughout these studies is that its power is covert. When given the opportunity to identify the most influential aspect of the crime report, participants (in all four studies that include a metaphoric frame) ignore the metaphor. Instead, they cite the crime statistics (which are the same in both conditions) as being influential in their reasoning. Together these studies suggest that unbeknownst to us, metaphors powerfully shape how we reason about social issues. Further, the studies help shed light on the mechanisms through which metaphors influence our reasoning.

Ethics Statement

The experiments reported here were done in accordance with the Declaration of Helsinki. Additionally, they followed the ethical requirements of the Stanford University institutional review board and complied with ethics guidelines set forth by the IRB recommendations. Participants were informed that their data would be treated anonymously and that they could terminate the experiment at any time without providing any reason. We received written informed consent from all participants before they participated in an experiment.

Participants

In Experiment 1, 485 students – 126 from Stanford University and 359 from the University of California, Merced – participated in the study as part of a course requirement. Experiments 2–5 were conducted online with participants recruited from Amazon's mechanical Turk (347, 312, 185, and 190, respectively). In exchange for participation in the study, people were paid $1.60 – consistent with a $10/hour pay rate since the study took 5 to 6 minutes to complete.

Gathering data from these various sub-populations allowed us to sample a broader cross-section of the general population. This is important since people's conceptions of social issues like crime are likely to differ as a function of factors like socioeconomic status and personal experience. This is particularly true of the sample that was recruited online, which was more diverse than that available at Stanford specifically or on college campuses generally [18] .

Running Experiments 2–5 online also afforded careful control over our sample population. We used Mechanical Turk's exclusion capabilities and tracked IP addresses to ensure that participants were not repeatedly sampled. We also restricted our study to Turkers with a 95% or better performance record to ensure that we were sampling high quality participants (“Requesters” have the opportunity to publicly give positive or negative feedback to their participants, which can then be used as a criterion for future “Requesters”). At the end of the online version of the study we asked participants to describe their language history, current geographic location, and provide some background information. We then restricted our analysis to residents of the United States who were native English speakers. The characteristics of our samples are detailed in the Results section below.

In each of the five experiments, participants were presented with a survey that included a short paragraph about crime in the fictional city of Addison and some follow-up questions. The survey differed subtly between experiments, but always contrasted a crime-as-virus framing with a crime-as-beast framing.

It should be noted that there are two somewhat different metaphorical frameworks that treat crime as an illness. In one, the community or population is seen as an organism, and crime is a disease that is developing inside that organism (e.g., “Violent crime is a cancer that eats away at the very heart of society.”). In another, the community is seen as individual agents and crime is a contagious disease that can be passed on from one person to another forming an epidemic. In this paper the stimuli did not strongly distinguish between these different varieties of crime as illness metaphors, but doing so would be an interesting extension of this work, as these metaphors suggest somewhat different implications for treating crime.

Experiment 1.

In the first experiment, participants were presented with one of two versions of the crime paragraph. The two versions of the paragraph differed only in the embedded metaphor: In one, crime was a beast; in the other, crime was a virus. The majority of the paragraph consisted of crime statistics, which were the same in both versions. Half of the participants were given the crime-as-beast version and half the crime-as-virus version. The paragraph read:

Crime is a {wild beast preying on/virus infecting} the city of Addison. The crime rate in the once peaceful city has steadily increased over the past three years. In fact, these days it seems that crime is {lurking in/plaguing} every neighborhood. In 2004, 46,177 crimes were reported compared to more than 55,000 reported in 2007. The rise in violent crime is particularly alarming. In 2004, there were 330 murders in the city, in 2007, there were over 500.

This report was followed up with two questions: 1) In your opinion what does Addison need to do to reduce crime? 2) Please underline the part of the report that was most influential in your decision. This question was aimed at discovering if participants explicitly noticed or made use of the metaphor.

Experiment 2.

The crime report used in the second experiment was similar, but not identical to the one used in Experiment 1. Importantly, it instantiated the beast or virus metaphor for crime with a single word. It read as follows:

Crime is a {beast/virus} ravaging the city of Addison. Five years ago Addison was in good shape, with no obvious vulnerabilities. Unfortunately, in the past five years the city's defense systems have weakened, and the city has succumbed to crime. Today, there are more than 55,000 criminal incidents a year - up by more than 10,000 per year. There is a worry that if the city does not regain its strength soon, even more serious problems may start to develop.

In Experiment 2, we asked three follow-up questions in the following order: 1) In your opinion what does Addison need to do to reduce crime? 2) What is the role of a police officer in Addison? 3) Please copy the part of the report that was most influential and paste it in the text area below. Questions one and two were free-response. Question three was copy and paste (participants were shown the report adjacent to an open text field and were asked to copy the portion of the report that was most influential in their reasoning and paste it into the open text field).

Experiment 3.

The design of Experiment 3 was similar to that of Experiment 2; however, before participants read the crime report, they were shown the word “beast” or the word “virus” and were asked to “list a synonym” for it. After completing this task, they were presented with the paragraph on crime in Addison on a separate screen. The crime report used in Experiment 3 was the same as the crime report for Experiment 2, except that it did not contain a virus or beast metaphor. The first sentence of the report read: “Crime is ravaging the city of Addison.” It was otherwise identical to the report from Experiment 2.

Experiment 4.

The crime report used in Experiment 4 was the same as the crime report used for Experiment 2. However, instead of asking the follow-up questions from Experiments 2 and 3, we asked participants to select one of four crime-related issues for further investigation – with the knowledge that this information should be used to help them make a more informed crime-reducing suggestion. The instructions read as follows: “Now imagine that Addison has consulted you about the crime problem. You have the resources to investigate one of the following four issues. Please select one from the list below.” The issues included: 1) the education system and availability of youth programs, 2) the economic system including the poverty level and employment rate, 3) the size and charge of the police force, and 4) the correctional facilities including the methods by which convicted criminals are punished.

Experiment 5.

The materials and task in Experiment 5 were identical to those of Experiment 4 except, instead of presenting the metaphor frame at the beginning of the report, we presented the metaphor frame at the end of the report, as shown below. All other aspects of the design were identical to Experiment 4. The paragraphs used were:

Five years ago Addison was in good shape, with no obvious vulnerabilities. Unfortunately, in the past five years the city's defense systems have weakened, and the city has succumbed to crime. Today, there are more than 55,000 criminal incidents a year - up by more than 10,000 per year. There is a worry that if the city does not regain its strength soon, even more serious problems may start to develop. Crime is a {beast/virus} ravaging the city of Addison.

In Experiment 1 the survey was included in a larger packet of questionnaires that were unrelated to this study.

In Experiments 2–5, each step of the experiment was presented on a separate screen. That is, the initial crime report was presented on a screen by itself. After participants read the report and clicked a button indicating they had finished reading it, the report disappeared and the first follow-up question appeared on a screen by itself. Similarly, each subsequent question was shown on a separate screen. On the final screen, participants were asked several background questions (e.g., What is the first language you learned to speak?).

Participants in Experiments 2–5 were explicitly instructed not to use the “back” button on their browser. If they did use the “back” button, the experimental session was terminated. This ensured that participants did not reread the crime report when they were later asked questions about it.

Experiment 1

In Experiment 1, we explored whether framing a crime problem with one of two contrasting metaphors for crime could systematically influence how people reasoned about the problem. Participants were presented with one of two versions of the crime paragraph (as detailed above) and asked a set of free response follow-up questions. Of particular interest, participants were asked how they would recommend solving Addison's crime problem.

Proposed solutions to the crime problem in Addison were coded into two categories in line with the results of the norming study described in the introduction: 1) diagnose/treat/inoculate, and 2) capture/enforce/punish. Responses were categorized as “diagnose/treat/inoculate” if they suggested investigating the underlying cause of the problem (e.g., “look for the root cause”) or suggested a particular social reform to treat or inoculate the community (e.g., fix the economy, improve education, provide healthcare). Responses were categorized as “capture/enforce/punish” if they focused on the police force or other methods of law enforcement (e.g., calling in the National Guard) or modifying the criminal justice system (e.g., instituting harsher penalties, building more jails). For brevity, we will refer to the “diagnose/treat/inoculate” category as “reform” and the “capture/enforce/punish” category as “enforce.”

Each participant's response was weighted equally – as a single point towards the analysis. For solutions that solely emphasized either reform or enforcement, the respective category was incremented by a point. Responses that exclusively emphasized one approach were the majority. Occasionally, however, participants listed both types of suggestions. In this case, if the response listed a disproportionate number of suggestions that were consistent with one approach (e.g., if the response listed three suggestions in line with reform and only one in line with enforcement, as in “investigate the root cause, institute new educational programs, create jobs, and hire more police”) then it was coded as a full point for the corresponding category. However, if the response equally emphasized both approaches, then the point was split between the categories such that each was incremented by .5.

Thirty of the 485 responses (6%) did not fit into either category. In every case this was because the response lacked a suggestion (e.g., “I don't know”, “I need more information”, “It should be addressed”). These data were omitted from analysis.

Participants' crime reducing suggestions were coded blindly by two coders. Cohen's kappa – a measure of inter-rater reliability – was .75 indicating good agreement between the coders ( p <.001). All disagreements between the coders were resolved between them before analyzing the data.

Overall, participants were more likely to emphasize enforcement strategies (65%) than reform (35%), χ 2  = 41.85, p <.001. However, as predicted, the solutions participants proposed to the crime problem in Addison differed systematically as a function of the metaphorical frame encountered in the crime report (see Fig. 1 ). Participants given the crime-as-beast metaphorical framing were more likely to suggest enforcement (74%) than participants given the crime-as-virus framing (56%), χ 2  = 13.94, p <.001. See Table 1 for response frequencies.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0016782.g001

Interestingly, when asked to identify the most influential aspect of the report, most participants ignored the metaphor. Only 15 participants (3%) identified the metaphoric frame as influential to their problem solving strategy. Removing these participants from the analysis did not affect the results (the proportion of responses that were congruent with the metaphor was not different in the two analyses, χ 2  = .0001, p  = .991). The vast majority of the participants identified the statistics in the crime report as being most influential in their decision – namely, the final three sentences of the paragraph that state the increasing crime and murder rate.

https://doi.org/10.1371/journal.pone.0016782.t001

Discussion.

In this experiment, we found that crime-reducing suggestions differed systematically as a function of the metaphor used to frame the crime problem. Participants who read that crime was a virus were more likely to propose treating the crime problem by investigating the root causes of the issue and instituting social reforms than participants who read that crime was a beast. Participants who read that crime was a beast were more likely to propose fighting back against the crime problem by hiring police officers and building jails – to catch and cage the criminals – than participants who read that crime was a virus.

Further, despite the clear influence of the metaphor, we found that participants generally identified the crime statistics, which were the same for both groups, and not the metaphor, as the most influential aspect of the report. These findings suggest that metaphors can influence how people conceptualize and in turn approach solving an important social issue, even if people don't explicitly perceive the metaphor as being especially influential.

Experiment 2

In Experiment 2, we made two substantive changes to the task to further test the role of metaphor in reasoning. First, we changed how the metaphoric frame was presented. In Experiment 1, the metaphoric frame was established several times and included vivid relational language. For example, crime was said to be either preying & lurking, or infecting & plaguing the community. These metaphorical verbs explicitly specified relations between crime and the community. Is specifying relations explicitly in this way necessary for people to make appropriate inferences, or might people be able to spontaneously extract the relevant relational inferences given a minimal metaphorical suggestion? Might a single carefully chosen and appropriately placed word be enough to instantiate a metaphorical frame and induce different reasoning strategies?

In Experiment 2 we tested this hypothesis by removing the relational verbs from the report. We replaced them with a single word metaphor that described crime as a “virus” or “beast” in the introductory sentence. The two conditions differed only in this one word, and otherwise included all the same information.

The second change we made was that we added an additional follow-up question: What is the role of a police officer in Addison? This question aimed to disambiguate the modal crime-reducing suggestion from Experiment 1, which was “increase the police force.” In that context, we interpreted the response (and close variants of it) as a suggestion for increased law enforcement and punishment. However, police officers do not just catch and punish criminals. They also serve as crime deterrents, educators, and role models and it is possible that some participants intended for the increased police presence to serve in this way. Including this question allowed these participants an opportunity to explicitly specify how they envisioned the increased police force impacting the community.

Participant characteristics.

We restricted our analysis of the initial sample of 347 Turkers to residents of the United States who were native English speakers. This left data from 253 participants for analysis (i.e., 94 participants were excluded – 27% of the initial dataset). Of these 253 participants, 157 were female and 96 were male. Their ages ranged from 18 to 66, with a mean age of 32 (and median age of 29). Eighty-two reported an affiliation to the Democratic Party, 57 reported an affiliation to the Republican Party, and 114 were Independent.

Crime-reducing suggestions were coded into two groups (reform and enforcement) as they were in Experiment 1. However, in Experiment 2 we coded one additional feature of this question: whether the participant exclusively suggested increasing the police force. For these responses, we planned to use the follow-up question about the role of a police officer in Addison to disambiguate whether the participant thought a police officer's primary role was as an instrument of social reform and prevention or an instrument of law enforcement and punishment.

Interpretations of the role of a police officer were coded into two groups that were analogous to the categories created for the first question: 1) crime deterrent, and 2) law enforcer and punisher. Interpretations that emphasized the police officer's role in preventing crime, educating youth, or serving as a role model in the community were coded as “crime deterrent.” Interpretations that emphasized the police officer's role in catching criminals, responding to crime reports, or punishing criminals were coded as “law enforcer and punisher.” As in Experiment 1, each response contributed one point to the analysis. This point either went entirely to one of the two categories or was split evenly between them.

Seven (3%) crime-reducing suggestions and 18 (7%) police officer interpretations were not coded. In every case this was because the response lacked a suggestion or interpretation and were eliminated from the analysis. It is possible that relatively more police officer interpretations fell into this category because the question was not prefaced with “In your opinion” (several responses to this question were a variant of “the report didn't say what the role of a police officer in Addison was”).

Answers to both of the free response questions were coded blindly by two coders. Inter-rater reliability was high for both: Cohen's kappa for crime-reducing suggestions was .86 ( p <.001); Cohen's kappa for interpretations of the role of a police officer was .72 ( p <.001). All disagreements between the coders were resolved between them before analyzing the data.

The results of Experiment 2 replicate our findings from Experiment 1. Participants were again overall more likely to suggest enforcement (62%) than reform (38%), χ 2  = 13.67, p<.01. However, the tendency towards enforcement was more pronounced among participants who read that crime was a beast (71%) than among participants who read that crime was a virus (54%), χ 2  = 6.50, p<.05. See Table 1 for response frequencies by condition.

Of the responses, 81 (31%) exclusively suggested increasing the police force. Disambiguating these responses by the participants' corresponding views of the role of a police officer in Addison further clarified the effect of the metaphor. Because “police” responses were previously coded as enforcement, disambiguating them created an overall shift to the reform category in both conditions, with a larger shift in the virus condition as predicted. With the “police” responses disambiguated, 37% of the responses advocated enforcement in the virus condition, and 59% advocated enforcement in the beast condition, χ 2  = 10.76, p<.01 (see Fig. 2 ).

The left panel displays results from Experiment 2 (with a one-word metaphor frame); the right panel displays results from Experiment 3 (in which a synonyms task preceded the non-metaphorically framed paragraph).

https://doi.org/10.1371/journal.pone.0016782.g002

Further, as in Experiment 1, participants did not explicitly report the metaphor as being influential in their reasoning. Only 18 of the 253 participants (7%) identified the metaphor as influential. Excluding participants who identified the metaphor as influential did not change the reported results (the proportion of responses that were congruent with the metaphor was not different in the two analyses, χ 2  = .01, p = .92).

Results of Experiment 2 replicate and extend the findings of Experiment 1. Manipulating the metaphor used to frame the issue of crime influenced how people approached solving the crime problem. When crime was framed as a virus, participants were more likely to suggest social reform. Alternatively, when crime was framed as a beast, participants were more likely to suggest law enforcement and punishment.

Remarkably, presenting an otherwise identical report with only one word different in the introductory frame (“Crime is a virus/beast ravaging the city of Addison”) yielded systematically different problem solving suggestions just as in Experiment 1. While in Experiment 1, the metaphoric frame was established using vivid verbs with rich relational meaning (e.g., crime was said to be either preying & lurking, or infecting & plaguing). The findings of Experiment 2 demonstrate that these relational elements need not be specified explicitly. People spontaneously instantiated the relevant relational inferences even given a single metaphorical noun in Experiment 2.

Further, in Experiment 2 we asked participants to provide their views on the role that a police officer should play in Addison. This afforded us a clearer interpretation of their crime-reducing suggestions and boosted our power to detect the influence of the metaphor.

Interestingly, despite the clear influence of the metaphor, we found that participants generally identified the crime statistics, which were the same for both groups, and not the metaphor, as the most influential aspect of the report.

Experiment 3

In Experiment 3 we tested whether the influence of the metaphor observed in the first two studies could have come about through simple spreading activation from lexical associates of the words “beast” and “virus.” Perhaps simply hearing a word like beast, even outside of the context of crime, would activate lexical associates like “hunting” and “caging”. These activated lexical associates might then color people's descriptions of how to solve the crime problem. To test for this possibility we dissociated the words “beast” and “virus” from the rest of the crime report in Experiment 3. Before reading the crime report, participants were asked to provide a synonym to the word “beast” or the word “virus” – thereby priming representations for a beast or a virus. They then read the same report about crime as in Experiment 2, but with the metaphorical word omitted (“Crime is ravaging the city of Addison”). Might a non-metaphorical lexical prime have the same effect as a metaphor?

Of the 312 Turkers that were initially sampled for Experiment 2, 76 (24%) were excluded because they did not live in the United States or because they were not native English speakers. This left data from 236 participants for analysis. Of these 236 participants, 136 were female and 100 were male. Their ages ranged from 18 to 81, with a mean age of 29 (and median age of 26). Seventy-six reported an affiliation to the Democratic Party, 48 reported an affiliation to the Republican Party, and 112 were Independent.

Answers to the free response questions were coded as they were in Experiment 2. Fifteen crime-reduction suggestions (6%) and 21 police officer interpretations (9%) did not fit into either category. In every case this was because the response lacked a suggestion or interpretation.

Answers to both of the free response questions were coded blindly by two coders. Inter-rater reliability was high for both: Cohen's kappa for crime-reducing suggestions was .87 ( p <.001); Cohen's kappa for interpretations of the role of a police officer was .84 ( p <.001). All disagreements between the coders were resolved between them before analyzing the data.

The synonyms that participants listed were analyzed to ensure that the lexical prime had the intended effect. Of the 124 participants in the crime-as-beast condition, all except one listed a synonym of “beast”. The modal response was “animal”, but others included “monster”, “mongrel”, “invader”, etc. The single respondent who did not list a synonym to “beast” instead wrote “I forget what a synonym is.” This participant's subsequent responses were omitted from the analyses reported below. Of the 112 participants in the crime-as-virus condition, all listed a synonym of virus. In this case, the modal response was “disease”, but others included “bug”, “cold”, “sickness”, “illness”, etc.

In Experiment 3, unlike Experiments 1 and 2, there was no difference in crime-reducing suggestions as a function of the condition – i.e., whether the participant listed a synonym to “virus” or “beast” before reading the crime paragraph did not affect what solutions they suggested to the crime problem. Overall, participants were significantly more likely to suggest enforcement or punishment (64%) than social reform (36%), χ 2  = 18.0, p <.001; however, there was no difference between participants who were lexically primed with “beast” (64% suggesting enforcement and punishment) versus those who were lexically primed with “virus” (65%), χ 2  = .001, p  = .99. See Table 1 for response frequencies by condition.

Further, disambiguating the responses that called for an increase to the police force did not differentiate the groups. Sixty-eight of the 235 responses (29%) were disambiguated. Of these, 29 (43%) interpreted the role of a police officer as a crime deterrent, 37 (54%) interpreted the role of a police officer as a law enforcer or punisher, and two responses could not be disambiguated. This disambiguation did not reveal a difference between conditions: Participants who were lexically primed with “virus” were no more likely to suggest enforcement (50%) than those who were lexically primed with “beast” (51%), χ 2  = .006, p  = .94 (see Fig. 2 ).

Comparing the results from Experiments 2 and 3 we find an interaction between the form in which the word “beast” or “virus” is presented (i.e., metaphor vs. lexical prime) and the extent to which crime-reducing suggestions are congruent with the prime. That is, we find that the metaphor in Experiment 2 was significantly more influential than the lexical prime in Experiment 3. To quantitatively compare the results of the two experiments we performed a chi-square contingency test as well as a set of logistic regressions. In Experiment 2, 61% of the responses were congruent with the metaphor (i.e., suggested “reform” when presented with crime-as-a-virus or suggested “enforcement” when presented with crime-as-a-beast), whereas only 50% of the responses in Experiment 3 were congruent with the lexical prime, χ 2  = 4.23, p <.05. Similarly, a logistic regression revealed that an interaction term for experiment X condition was a significant predictor of people's crime-fighting suggestions: a model that included the three predictors (experiment, condition, and the interaction term) was significantly better than a model with two predictors (omitting the interaction term), χ 2 (1, 459) = 5.85, p< .05.

In Experiment 3 we tested whether the influence of the metaphor observed in the first two studies could have come about through simple spreading activation from lexical associates of the words “beast” and “virus.” We dissociated the words “beast” and “virus” from the story, so that they could act as non-metaphorical lexical primes. These disconnected lexical primes did not yield differences in people's crime-fighting suggestions. These results suggest that metaphors act as more than just isolated words – their power appears to come from participating in elaborated knowledge structures.

Additionally, the results of Experiment 3 shed some light on this population's baseline preference for reducing crime. That is, in Experiment 2 it might have been the case that participants had a general preference for reducing crime through enforcement and that it was the crime-as-virus frame alone that shifted peoples' responses. The results of Experiment 3, however, suggest that the population does not seem to favor either of the two crime-reducing suggestions absent a metaphoric frame and that both frames are influential.

Experiment 4

In Experiment 4 we tested whether the influence of the metaphor would persevere even if people were able to select responses from a full set of options. One possibility is that a metaphorical frame affects what kind of solution comes to mind easiest. However, when faced with a complete set of options, people may realize they had neglected to attend to other alternatives and no longer show the influence of the metaphor. For example, a participant in the “beast” frame may not have spontaneously thought to address underlying problems in the economy or education. However, if these are made explicitly available as response options, the participant may recognize them as good ideas and may re-bound from the metaphorical framing. To test for this, in Experiment 4, we presented participants with a list of four possible approaches to the crime problem and asked them to choose one. These included two options that were more consistent with social reform (education, economy) and two options that were more consistent with enforcement and punishment (police, jails).

Rather than asking participants to make a crime-reducing suggestion as in previous studies, the task in Experiment 4 was to select an area to investigate further (in preparation to making a crime-fighting suggestion). This aspect of the experiment was designed to test whether metaphors can affect not only how people propose solving the problem of crime, but also how they go about gathering information for future problem solving. If participants seek out information that is likely to confirm the initial bias suggested by the metaphor, this may be a mechanism for metaphors to iteratively amass long-term effects on people's reasoning (as people seek out more and more confirming evidence).

Of the 185 Turkers who participated in Experiment 4, seven (4%) were excluded because they did not live in the United States or because they were non-native English speakers. This left data from 178 participants for analysis. Of these 178 participants, 89 were female and 89 were male. Their ages ranged from 18 to 70, with a mean age of 31 (and median age of 28). Seventy-eight reported an affiliation to the Democratic Party, 28 reported an affiliation to the Republican Party, and 72 were Independent.

Choosing to gather additional information about the education system or economic system was coded as a social reform category of response; gathering additional information about the police force or criminal justice system was coded as an enforcement and punishment category of response.

Results of Experiment 4 replicate the effects of metaphorical frames found in Experiments 1 and 2. Participants who were presented with the crime-as-a-beast metaphor were more likely to gather additional information about the city's criminal justice system (40%) than participants who were presented with the crime-as-a-virus metaphor (22%), χ 2  = 5.72, p <.05. See Table 1 for response frequencies by condition.

As we saw in Experiments 1 and 2, when given the opportunity to identify the most influential aspect of the report, the vast majority ignored the metaphor. Only 27 participants (15%) reported that the metaphor influenced their decision. Eliminating these participants from the analysis does not change the results (the proportion of responses that were congruent with the metaphor was not different in the two analyses, χ 2  = .003, p  = .96).

In Experiment 4 we found that the effect of metaphorical framing persists even when the list of all possible approaches to solving crime is explicitly presented. Laying out four possible approaches to crime shifted the overall likelihood that people wanted to pursue social reform. It seems that explicitly seeing the space of possible responses makes people more likely to attempt reducing crime through reform than enforcement. However, we still found that peoples' responses were influenced by the frame that they read. Additionally, the results of Experiment 4 reveal that the metaphorical frame influences how people go about gathering information for future problem solving. People tended to seek additional information about the city that confirmed their initial (metaphor-induced) suspicion about how to solve crime.

Experiment 5

In Experiment 5 we investigated the time-course of how metaphors influence people's construal of and reasoning about problems. One possibility is that metaphors influence reasoning by instantiating a knowledge frame that structures subsequent information. After being exposed to the metaphor, participants may assimilate all further information they receive into this knowledge structure, instantiating any ambiguous information in a way that would be consistent with the metaphor. For example, words like “vulnerabilities”, “defense”, “weakened” may take on different meanings depending on whether they are understood in the context of viruses or beasts [13] , [19] . If this is the case, if metaphors actively coerce incoming information, then metaphors should have the most impact when they are presented early, such that their impact can accumulate in the course of assimilating further information.

Alternatively, if metaphors simply activate a fossilized package of ideas and do not encourage the kind of assimilation process described above, then they should be most effective when they are presented late in the narrative, as close to when people are asked to reason about a solution as possible. This way, the memory of the metaphor should be fresh and any knowledge activated by it should have the best chance to influence reasoning. In Experiment 5, we repeated the design of Experiment 4, but moved the metaphorical frame so that instead of being the first sentence in the crime report it was the last.

As in Experiment 4, choosing to gather additional information about the education system or economic system was coded as a social reform category of response; gathering additional information about the police force or criminal justice system was coded as an enforcement and punishment category of response.

As in Experiment 4, participants in Experiment 5 were overall more likely to gather information relating to the city's social situation (67%) than the criminal justice system (33%), χ 2  = 19.55, p <.001.

However, unlike Experiment 4, there was no effect of the metaphorical frame. Participants who were presented with the crime-as-a-beast metaphor were about equally likely to gather additional information about the city's social situation (69%) as participants who were presented with the crime-as-a-virus metaphor (64%), χ 2  = .29, p  = .59 (see Fig. 3 ). See Table 1 for response frequencies by condition.

The left panel displays results from Experiment 4 (with a one-word metaphor frame at the beginning of the report); the right panel displays results from Experiment 5 (with the same one-word frame but at the end of the report).

https://doi.org/10.1371/journal.pone.0016782.g003

This pattern was significantly different from the effects found in Experiment 4, χ 2  = 5.45, p <.05. That is, significantly more participants were influenced by the metaphor when it was presented at the beginning of the report (Experiment 4) than at the end of the report (Experiment 5). This conclusion is also supported by a logistic regression, which revealed that an interaction term for experiment X condition was a significant predictor of people's crime-fighting suggestions: a model that included the three predictors (experiment, condition, and the interaction term) was significantly better than a model with two predictors (omitting the interaction term), χ 2 (1, 346) = 5.34, p< .05.

As we saw in the previous experiments, when given the opportunity to identify the most influential aspect of the report, the vast majority ignored the metaphor. Only 18 participants (10%) reported that the metaphor influenced their decision.

In Experiment 5 we investigated whether when a metaphor is introduced affects the metaphor's influence. Experiment 5 repeated the design of Experiment 4, but we moved the metaphorical frame so that instead of being the first sentence in the crime report it was the last. Unlike the results of Experiment 4, this late metaphorical framing had no effect on people's crime-related information foraging. These findings suggest that metaphors can gain power by coercing further incoming information to fit with the relational structure suggested by the metaphor.

These results are particularly striking since in Experiment 5, the metaphorical frame appears in much closer proximity to the measure of interest. It would have been reasonable to predict that a metaphorical frame that is more fresh in mind should have the largest effect. Instead, the way a metaphorical frame is integrated into the narrative appears to be more important. This finding also helps allay a possible worry about the findings in Experiment 3. In Experiment 3, we moved the words “virus” and ”beast” out of the crime story, and asked participants to generate synonyms to these words before they read about crime. When the words appeared in this way as disconnected lexical primes, they had no influence over people's crime-fighting suggestions. Of course, one possibility is simply that taking the words out of the narrative also made them more distant in time from the measure of interest. Results of Experiment 5 suggest that it is integration at the right point in the narrative rather than simple temporal distance that modulates our effects. In Experiment 5, the words “virus” and “beast” occurred immediately prior to the measure of interest, and yet had no effect.

In five experiments we investigated the role of metaphor in guiding how people reason about the complex problem of crime. We found that metaphors exert an influence over people's reasoning by instantiating frame-consistent knowledge structures, and inviting structurally-consistent inferences. Further, when asked to seek out more information to inform their decisions, we found that people chose information that was likely to confirm and elaborate the bias suggested by the metaphor – an effect that persisted even when people were presented with a full set of possible solutions.

Our results suggest that even fleeting and seemingly unnoticed metaphors in natural language can instantiate complex knowledge structures and influence people's reasoning in a way that is similar to the role that schemas [20] , [21] , scripts [22] , [23] , and frames [24] have been argued to play in reasoning and memory [13] , [25] – [27] . That is, the metaphors provided our participants with a structured framework for understanding crime in Addison, influenced the inferences that they made about the crime problem, and suggested different causal interventions for solving the problem. This was true even though the metaphors themselves did not strike our participants as particularly influential.

Consistent with previous work on meaning instantiation, we find that the metaphors were most effective when they were presented early in the narrative and were then able to help organize and coerce further incoming information. For example, Bransford and Johnson demonstrate that a procedural description of washing clothes was understood and remembered best when participants knew the topic of the passage before they heard the description [13] . When the topic was given at the end of the passage or not at all, participants reported being unable to make sense of what they had heard and were able to recall few details of the description on a memory test. While the crime passage we used was clearly not as ambiguous as the procedural description of washing clothes used by Bransford and Johnson, it did contain many words and phrases that would likely be interpreted differently in the different contexts represented by the metaphoric frames. For instance, in the context of an attacking beast the meaning of the words “vulnerable” and “defense system” may be different from what the same words would be taken to mean in the context of a spreading virus. Previous work has demonstrated that contextual cues can strongly influence how people interpret seemingly unambiguous text [19] , [28] – [29] .

A further question is how such knowledge structures for thinking about crime emerge? How do people build virus-like or beast-like representations of crime and what is the role of linguistic metaphor in encouraging the construction of such knowledge structures? One potential mechanism is offered by work in analogical reasoning [6] , [30] – [35] . For example, Bowdle & Gentner suggest that metaphors when first encountered are processed as analogies or structural alignments [35] . When we first hear about crime described as a beast, for example, we may carry out comparisons to discover any alignable similarities between crime and beasts. If such similarities are discovered, they can license the transfer of inferences from one domain to the other, and the most striking or stable structural similarities can be highlighted and stored in memory. With exposure to the system of “beast” metaphors, an elaborated knowledge structure can emerge for thinking about crime that mirrors in important relational structure the representations we have about the behavior of wild beasts. Through analogical transfer in this way, systems of metaphors in language can encourage the creation of systems of knowledge in a wide range of domains. Our reasoning about many complex domains then can be mediated through these patchworks of analogically-created representations.

A final question is how strong the influence of metaphorical framing really is? Focusing on a real-world social issue like crime allows us to compare the effects of metaphor we observe in the lab with the opinion differences that exist naturally in the population. People with different political affiliations hold different opinions on how to address societal problems like crime. How do the differences we find between metaphorical conditions compare to those between Democrats and Republicans, for example?

At the end of Experiments 2–5, we asked participants to report their political affiliation (Democrat, Independent, or Republican) and their gender. We found a predictable relationship between political affiliation and the tendency to emphasize enforcement in one's response. Across the four experiments, 48% of responses from Republicans emphasized enforcement whereas only 40% of responses from Democrats and Independents emphasized enforcement (data from Democrats and Independents did not differ from one another and so were collapsed). A logistic regression revealed political affiliation to be a significant predictor of people's crime-fighting suggestions: comparing a model with political affiliation included as a predictor to a constant-only model was statistically significant, χ 2 (1, 839) = 3.98, p< .05. We also found systematic differences by gender: 46% of responses from men and 38% of responses from women suggested enforcement. Comparing a logistic regression model with gender included as a predictor to a constant-only model was statistically significant, χ 2 (1, 839) = 5.389, p< .05.

Impressively the differences in opinion generated by the metaphorical frames were larger than those that exist between Democrats and Republicans, or between men and women. Metaphorical frames caused shifts of 18–22% in enforcement responses in Experiments 2 and 4. Differences between people of different political affiliations or between the two genders were 8–9%. To statistically compare the strength of these different predictors, we fit a set of logistic regression models for data from Experiments 2 and 4. We found that a model fit with a predictor for metaphor frame was significantly better than a constant-only model, χ 2 (1, 839) = 17.35, p< .001; however, including a predictor for gender, χ 2 (1, 839) = 0.013, ns , or political affiliation, χ 2 (1, 839) = 2.06, ns , or both, χ 2 (3, 839) = 3.03, ns , did not improve the model significantly. This analysis reveals a striking effect of metaphor as measured against real-world differences in opinion that exist in the population and impact policy-making.

Interestingly, we found that self-identified Republicans were also less likely to be influenced by the metaphors than were Democrats and Independents. Looking at data from Experiments 2 and 4 we find that 63% of the responses from Democrats and Independents are congruent with the metaphorical frame, whereas only 49% of those from Republicans were congruent with the metaphor. A logistic regression revealed that political affiliation was indeed a significant predictor of congruence with the metaphorical frame: comparing a model with political affiliation as a predictor against a constant-only model was statistically significant, χ 2 (1, 839) = 5.46, p<.05 . These results may be consistent with previous analyses showing a difference in openness between people of different political affiliations [36] . Men and women were equally influenced by the metaphorical frames.

The studies presented in this paper demonstrate that even minimal (one-word) metaphors can significantly shift people's representations and reasoning about important real-world domains. These findings suggest that people don't have a single integrated representation of complex issues like crime, but rather rely on a patchwork of (sometimes disconnected or inconsistent) representations and can (without realizing it) dynamically shift between them when cued in context.

Metaphor is incredibly pervasive in everyday discourse. By some estimates, English speakers produce one unique metaphor for every 25 words that they utter [37] . Metaphor is clearly not just an ornamental flourish, but a fundamental part of the language system [28] , [38] . This is particularly true in discussions of social policy [5] , [39] – [40] , where it often seems impossible to “literally” discuss immigration, the economy, or crime. If metaphors routinely influence how we make inferences and gather information about the social problems that confront us, then the metaphors in our linguistic system may be offering a unique window onto how we construct knowledge and reason about complex issues.

Conclusions

The way we talk about complex and abstract ideas is suffused with metaphor. In five experiments, we have explored how these metaphors influence the way that we reason about complex issues and forage for further information about them. We find that metaphors can have a powerful influence over how people attempt to solve complex problems and how they gather more information to make “well-informed” decisions. Our findings shed further light on the mechanisms through which metaphors exert their influence, by instantiating frame-consistent knowledge structures, and inviting structurally-consistent inferences. Interestingly, the influence of the metaphorical framing is covert: people do not recognize metaphors as an influential aspect in their decisions. Finally, the influence of metaphor we find is strong: different metaphorical frames created differences in opinion as big or bigger than those between Democrats and Republicans.

Acknowledgments

The authors would like to thank Jay McClelland, Caitlin Fausey, Alexia Toskos, Steve Flusberg, Tania Henetz, and members of the Cognation Lab for inspiration and helpful, critical discussion of the content of this paper and the issues it addresses. We would also like to than Otto Murphy and Nguyen Ngo for their help in coding responses. This material is based on work supported under a Stanford Graduate Fellowship to the first author.

Author Contributions

Conceived and designed the experiments: PHT LB. Performed the experiments: PHT. Analyzed the data: PHT. Contributed reagents/materials/analysis tools: PHT. Wrote the manuscript: PHT LB.

• 1. Federal Bureau of Investigation Uniform crime report (2008) - http://www.fbi.gov/ucr/cius2008/data/table_01.html .
• 2. Walmsley R (2009) World prison population list (eighth edition). London, UK: King's College of London.
• View Article
• Google Scholar
• 4. Lakoff G, Johnson M (1980) Metaphors We Live By. Chicago: Univ. of Chicago Press.
• 5. Lakoff G (2002) Moral Politics: How Liberals and Conservatives Think. Chicago: University of Chicago Press.
• 6. Gentner D, Gentner D (1983) Flowing waters and teeming crowds: mental models of electricity. In: Gentner D, Stevens A, editors. Mental Models. Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers.
• 7. Gibbs R (1994) The poetics of mind: Figurative thought, language, and understanding. New York: Cambridge University Press.
• 14. Pinker S (2007) The stuff of thought: Language as a window in human nature. New York, NY: Penguin Books.
• 20. Rumelhart DE (1975) Notes on a schema for stories. In: Bobrow DG, Collins AM, editors. Representaion and understanding: Studies in cognitive science. New York: Academic Press.
• 21. Rumelhart DE, Ortony A (1977) The representation of knowledge in memory. In: Anderson RC, Spiro RJ, Montague WE, editors. Schooling and the acquisition of knowledge. Hillsdale, N.J.: Lawrence Erlbaum Associates.
• 22. Schank RC, Abelson RP (1975) Scripts, plans and knowledge. Advance Papers of the Fourth International Joint Conference on Artificial Intelligence. Tbilisi, Georgia: USSR. pp. 151–157.
• 24. Minsky M (1975) A framework for representing knowledge. In: Winston PH, editor. The psychology of computer vision. New York: McGraw-Hill.
• 25. Bartlett FC (1932) Remembering: A study in experimental and social psychology. Cambridge, England: Cambridge Univ. Press.
• 28. Rumelhart DE (1979) Some problems with the notion of literal meanings. In: Ortony A, editor. Metaphor and Thought. Cambridge, UK: Univ. Press.
• 34. Gentner D, Bowdle B, Wolff P, Boronat C (2001) Metaphor is like analogy. In: Centner D, Holyoak KJ, Kokinov BN, editors. The analogical mind: Perspectives from cognitive science. Cambridge, MA: MIT Press. pp. 199–253.
• 39. Lakoff G (2008) The Political Mind: Why you can't understand 21 st century politics with an 18 th century brain. New York, NY: Penguin Books.
• 40. Schön D (1993) Generative metaphor: A perspective on problem-setting in social policy. In: Ortony A, editor. Metaphor and Thought. Cambridge, UK: Univ. Press.

• Frontiers in Psychology
• Psychology of Language
• Research Topics

Metaphor Studies: Theories, Methods, Approaches, and Future Perspectives

Total Downloads

Total Views and Downloads

About this Research Topic

Metaphor has been studied as a linguistic, conceptual, and communicative phenomenon in a wide range of disciplines, including linguistics, literary studies, philosophy, education, political science, media studies, communication science, psychology, and neurosciences. While the Cognitive Theory of Metaphor has ...

Keywords : Metaphor, figurative language, metaphorical thought, abstract conceptualization, embodied metaphors, metaphor in discourse, metaphor variation

Important Note : All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.

Topic Editors

Topic coordinators, recent articles, submission deadlines, participating journals.

Manuscripts can be submitted to this Research Topic via the following journals:

total views

• Demographics

No records found

total views article views downloads topic views

Top countries

Top referring sites, about frontiers research topics.

With their unique mixes of varied contributions from Original Research to Review Articles, Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author.

Metaphor Analysis

• First Online: 02 January 2023

Cite this chapter

• Benjamin Kutsyuruba 4 &
• Jodi Basch 4  

Part of the book series: Springer Texts in Education ((SPTE))

3843 Accesses

This chapter describes the metaphor analysis approach. Metaphorical language, as a form of interpretation of meaning, entails the use of figures of speech, like metaphors, analogy, simile, or synecdoche, to make implicit comparisons where a word or phrase that is ordinarily used in one domain is applied in another domain. By analyzing metaphorical language, researchers can search for metaphors in a variety of texts and derive meaning from them. This chapter outlines the brief history, purpose, and components of metaphor analysis, provides an outline of its process, strengths and limitations, and application, and offers further readings, resources, and suggestions for student engagement activities.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Available as EPUB and PDF
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Chenail, R. J. (2012). Conducting qualitative data analysis: Qualitative data analysis as a metaphoric process. The Qualitative Report, 17 (1), 248–253. http://www.nova.edu/ssss/QR/QR17-1/chenail-metaphor.pdf

Ignatow, G., & Mihalcea, R. (2017). Text mining: A guidebook for the social sciences . SAGE.

Google Scholar  

Kövecses, Z. (2017). Conceptual metaphor theory. In E. Semino & Z. Demjén (Eds.), Routledge Handbook of Metaphor (pp. 13–27). Routledge.

Lakoff, G., & Johnson, M. (1980). Metaphors we live by . University of Chicago Press.

Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh: The embodied mind and its challenge to Western thought . Basic Books.

Lakoff, G., & Johnson, M. (2003). Metaphors we live by (2nd ed.). University of Chicago Press.

Pitcher, R. (2013). Using metaphor analysis: MIP and beyond. The Qualitative Report, 18 (34), 1–8.

Pragglejaz Group. (2007). MIP: A method for identifying metaphorically used words in discourse. Metaphor and Symbol, 22 (1), 1–39

Schmitt, R. (2005). Systematic metaphor analysis as a method of qualitative research. The Qualitative Report, 10 (2), 358–394.

Steger, T. (2016). The stories metaphors tell: Metaphors as a tool to decipher tacit aspects in narratives. Field Methods, 19 (1), 3–23. https://doi.org/10.1177/1525822x06292788

Article   Google Scholar  

Additional Reading

Armstrong, S. L., Davis, H. S., & Paulson, E. J. (2011). The subjectivity problem: Improving triangulation approaches in metaphor analysis studies. International Journal of Qualitative Methods, 10 (2), 151–163.

Cameron, L., & Maslen, R. (2010). Metaphor analysis: Research practice in applied linguistics, social sciences and the humanities. Equinox.

Fábián, G. (2013). The application of improved metaphor analysis in education research. Procedia — Social and Behavioral Sciences, 93 (2013), 1025–1029. https://doi.org/10.1016/j.sbspro.2013.09.323

Jensen, D. (2006). Metaphors as a bridge to understanding educational and social contexts. International Journal of Qualitative Methods, 5 (1), 36–54. https://doi.org/10.1177/160940690600500104

Kimmel, M. (2012). Optimizing the analysis of metaphor in discourse: How to make the most of qualitative software and find a good research design. Review of Cognitive Linguistics, 10 (1), 1–48. https://doi.org/10.1075/rcl.10.1.01kim

Miller, S. I., & Fredericks, M. (1988). Uses of metaphor: A qualitative case study. International Journal of Qualitative Studies in Education, 1 (3), 263–272. https://doi.org/10.1080/0951839900040104c

Todd, Z., & Harrison, S. J. (2008). Metaphor analysis. In S. N. Hesse-Biber & P. Leavy (Eds.), Handbook of emergent methods (pp. 497–493). Guildford.

Online Resources

The art of a metaphor—Jane Hirshfield (5 ½ minutes): https://www.youtube.com/watch?v=A0edKgL9EgM

Metaphors we live by: George Lakoff and Mark Johnson (12 minutes) https://www.youtube.com/watch?v=lYcQcwUfo8c&t=8s

Qualitative data analysis—Closet metaphor (14 ½ minutes) https://www.youtube.com/watch?v=j1wj_BbK5Ec

200 short and sweet metaphor examples https://literarydevices.net/a-huge-list-of-short-metaphor-examples/

Download references

Author information

Authors and affiliations.

Queen’s University, Kingston, Canada

Benjamin Kutsyuruba & Jodi Basch

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Benjamin Kutsyuruba .

Editor information

Editors and affiliations.

Department of Educational Administration, College of Education, University of Saskatchewan, Saskatoon, SK, Canada

Janet Mola Okoko

Scott Tunison

Department of Educational Administration, University of Saskatchewan, Saskatoon, SK, Canada

Keith D. Walker

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

Kutsyuruba, B., Basch, J. (2023). Metaphor Analysis. In: Okoko, J.M., Tunison, S., Walker, K.D. (eds) Varieties of Qualitative Research Methods. Springer Texts in Education. Springer, Cham. https://doi.org/10.1007/978-3-031-04394-9_50

Download citation

DOI : https://doi.org/10.1007/978-3-031-04394-9_50

Published : 02 January 2023

Publisher Name : Springer, Cham

Print ISBN : 978-3-031-04396-3

Online ISBN : 978-3-031-04394-9

eBook Packages : Education Education (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

Get science-backed answers as you write with Paperpal's Research feature

Using Metaphors in Academic Writing

Have you ever wanted to translate formidable, and sometimes tedious, academic content into one that is easily comprehensible and captivating? Academics are often told that the language of science is formal, precise and descriptive with no space for the abstract. However, using metaphors in your academic writing could be helpful if used to explain complex scientific concepts. Just remember not to be cautious and exercise restraint when using different types of metaphors or it could make your academic writing seem unprofessional.

What is a metaphor?

A metaphor is defined as a figure of speech in which a word or phrase denoting one kind of object or idea is used in place of another to suggest a likeness or analogy between them. (Merriam-Webster, 2022). Derived from the Greek word ‘metapherein,’ which means ‘to transfer,’ metaphors transfer the meaning of one word to another to encourage a feeling. For example, by writing ‘ All the world’s a stage,’ Shakespeare creates a powerful imagery of ideas through transference. By bringing life to words, metaphors add value to writing and are a great addition to a writer’s toolkit.

Difference between similes and metaphors and analogies

When you’re writing in English, you should know the difference between similes and metaphors and analogies. While these are similar in terms of purpose, i.e., comparing two things, they are different in how they are used. A simile is explicit about the comparison, while a metaphor simply points to the similarities between two things, and an analogy seeks to use comparisons to explain a concept.

This could be confusing, however, there are simple ways to detect the differences between similes and metaphors and analogies. You can identify a simile by looking for the use of words ‘like’ , ‘as’, for example, ‘Life is like a box of chocolates.’ On the other hand, metaphors are more rhetorical and not so literal, for example, ‘The news was music to her ears.’ An analogy is more complex and seeks to point out the similarity in two things to explain a point, for example, ‘Finding the right dress is like finding a needle in a haystack.’

Types of metaphors

There are several different types of metaphors in the English language, here are some of the most common variations.

• Standard metaphor: A standard metaphor directly compares two unrelated items. For instance, by drawing a link between things and feelings, we’ve been able to convey the importance of laughter in this example of a metaphor: Laughter is the best medicine.
• Implied metaphor: This type of metaphor implies comparison without mentioning one of the things being compared. Take this example, where the coach’s voice is implied to be as loud as thunder: “Don’t give up!” thundered the coach from the side lines.
• Visual metaphor: This type of metaphor compares abstract objects or ideas that are difficult to imagine to a visual image that is easily identifiable; providing the former with a pictorial identity. This type of metaphor is most widely used in advertisements. For example, for the phrase ‘ The Earth is melting’ , the visual metaphor used to signal global warming is a melting ice cream.
• Extended metaphor: This type of metaphor extends the comparison throughout an article, document, or stanza. For example, when poet Emily Dickinson wrote “Hope” is the thing with feathers, she used feathers as a metaphor to compare hope to a bird with wings.
• Grammatical metaphors : Also known as nominalization, this type of metaphor rewrites verbs or adjectives as nouns. It’s most commonly used in academic and scientific texts as a way to separate spoken and written language, remove personal pronouns, and write in a concise manner. For instance, ‘ Millions of men, women and children starved to death in the 1943 Bengal Famine as a direct result of Churchill’s policies.’ This can be rephrased as ‘British policies led to the 1943 Bengal Famine, impacting the country’s people and politics for decades.’

Using metaphors in academic writing

Scholars pride themselves on creating research papers that are factually correct and precise, and metaphors may be perceived to detract from this. However, using metaphors may be a great way to explain scientific and technical concepts to readers, who may not know as much about the subject. While metaphors can add to formal academic writing and make it more engaging, it’s important to find a balance. Here are some tips to keep in mind when using metaphors in academic writing:

• Don’t use metaphors as the foundation of your academic content, use them instead to support your argument and drive home a point.
• Choose your metaphors carefully taking into account your primary audience; using figures of speech specific to any one region can introduce confusion instead of clarity.
• Use metaphors wisely and only when needed so not to distract the reader. They should flow naturally and enhance the content rather than detract from the point.

Metaphors are a nifty way to create engaging content even for academic writers. Greek philosopher Aristotle once wrote, “The greatest thing by far is to be a master of metaphor; it is the one thing that cannot be learnt from others.” So get ready to wield that pen and reach for the stars!

Paperpal is a comprehensive AI writing toolkit that helps students and researchers achieve 2x the writing in half the time. It leverages 21+ years of STM experience and insights from millions of research articles to provide in-depth academic writing, language editing, and submission readiness support to help you write better, faster.  

Get accurate academic translations, rewriting support, grammar checks, vocabulary suggestions, and generative AI assistance that delivers human precision at machine speed. Try for free or upgrade to Paperpal Prime starting at US$19 a month to access premium features, including consistency, plagiarism, and 30+ submission readiness checks to help you succeed.  

Experience the future of academic writing – Sign up to Paperpal and start writing for free!  

Related Reads:

• Quotation Marks: When and How to Use Quotations in Academic Writing
• Week vs. Weak: Bringing Out the Distinction
• Center vs. Centre: How to Differentiate Between The Two Words

Paraphrasing in Academic Writing: Answering Top Author Queries

How and when to use active or passive voice in research papers, transition phrases in academic writing, you may also like, how to use ai to enhance your college..., how to use paperpal to generate emails &..., word choice problems: how to use the right..., how to paraphrase research papers effectively, 4 types of transition words for research papers , sentence length: how to improve your research paper..., navigating language precision: complementary vs. complimentary, climatic vs. climactic: difference and examples, language and grammar rules for academic writing.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Review Article
• Open access
• Published: 09 December 2023

Research on metaphor processing during the past five decades: a bibliometric analysis

• Zhibin Peng 1 &
• Omid Khatin-Zadeh 2 , 3  

Humanities and Social Sciences Communications volume  10 , Article number:  928 ( 2023 ) Cite this article

1366 Accesses

1 Altmetric

Metrics details

• Language and linguistics

Metaphor processing has been the subject of extensive research over the past five decades. A systematic review of metaphor processing publications through bibliometric tools can provide a clear overview of research on metaphor processing. In this study, we used the CiteSpace bibliometric tool to conduct a systematic review of publications related to metaphor processing. A total of 3271 works published and indexed in the Web of Science (WoS) were gathered. These works had been published between 1970 and 2022. We analyzed the co-citations of these works by CiteSpace to identify the most influential publications in metaphor processing research. A co-occurrence term analysis was done to identify dominant topics in this area of research. The results of this analysis showed that Language, comprehension, metaphor, figurative language , and context were the most frequent keywords. The most prominent clusters were students, figurative language, right hemisphere, embodied cognition, comprehension, N400 , and anger . Based on the results of this analysis, we suggest that task properties such as response format and linguistic features should be carefully taken into account in future studies on metaphor processing.

Similar content being viewed by others

Anger is eliminated with the disposal of a paper written because of provocation

Yuta Kanaya & Nobuyuki Kawai

Microdosing with psilocybin mushrooms: a double-blind placebo-controlled study

Federico Cavanna, Stephanie Muller, … Enzo Tagliazucchi

Song lyrics have become simpler and more repetitive over the last five decades

Emilia Parada-Cabaleiro, Maximilian Mayerl, … Eva Zangerle

Introduction

How people understand and produce metaphors has long aroused the interest of scholars from various disciplines such as philosophy, linguistics, and psychology. From the 1970s, scholars began to study the processing of metaphors through experiments. Throughout the past five decades, a large body of experimental and theoretical works on metaphor have been produced, and many journals have started to publish papers related to metaphor. During this period, many researchers in neurolinguistics and psycholinguistics published their works on metaphor. These works have fundamentally changed the ways that researchers have been studying metaphors. This is particularly the case with research on metaphor processing. According to a study conducted by Han et al. ( 2022 ), research on metaphor processing has been the most active area of research on metaphor.

Metaphor processing research is an interdisciplinary area of study on metaphor that involves linguistics, psychology, and neuroscience. A large number of works on metaphor processing have been published in recent years, including reviews directed at selected subtopics. For instance, Rai and Chakraverty ( 2020 ) provided a systematic review of computational models and approaches to metaphor comprehension. This systemic review presented a concise yet representative picture of computational metaphor processing. In a related work, Kertész, Rákosi, and Csatár ( 2012 ) presented a review that was focused on the data, problems, heuristics, and results in cognitive research on metaphor. Some works have presented comprehensive reviews of studies conducted on metaphor comprehension in non-typical populations. For example, Morsanyi et al. ( 2020 ) conducted a systematic review and meta-analysis of metaphor processing in autism. Kalandadze et al. ( 2019 ) also presented a systematic review and meta-analysis of studies on metaphor comprehension in individuals suffering from autism. This review specifically focused on task properties. However, among these works, except for a review paper published by Holyoak and Stamenković ( 2018 ), no other publication has specifically focused on theories and evidence related to metaphor processing. Furthermore, the past review papers have been primarily based on subjective judgment rather than bibliometric tools. Therefore, a systematic review conducted by bibliometric tools can shed new light on our understanding of metaphor processing research. In the literature of the field, we found just two works on bibliometrics of conceptual metaphor research (Han et al., 2022 ; Zhao et al., 2023 ). These two works have presented bibliometric assessments of published works on conceptual metaphor theory. However, they are only about conceptual metaphors in general. To fill this gap in the literature of the field, we used CiteSpace to present a systematic review of studies on metaphor processing.

CiteSpace is a bibliometric analysis tool that can provide an exhaustive account of research in any area over a certain period of time. In this way, it can suggest some directions for future research. Compared to those reviews relying on subjective judgment, a review conducted by CiteSpace can help us navigate through the key documents, research fields, and dominant topics in metaphor processing. Importantly, the results of such analysis can be presented in the form of easily understandable diagrams. We intended to identify the most productive and influential journals, authors, and institutions in the field of metaphor processing. Also, we intended to identify the most influential documents, active research areas, and dominant topics in metaphor processing research. Specifically, by analyzing the co-occurrence of keywords associated with metaphor processing, we aimed to depict a cluster picture of related keywords and dominant topics in this area of research. In this way, we intended to answer the following research questions:

Q1: What are the active research areas and dominant topics in metaphor processing research?

Q2: Is it possible to use a cluster picture of related keywords and research topics to identify research features that play a critical role in studies on metaphor processing?

We hypothesized that a cluster picture of related keywords and research topics in metaphor processing can be used to identify critical research properties that can be taken into account in future studies on metaphor processing.

Methodology

Data collection.

As the study was focused on metaphor processing, we collected and analyzed the published documents by conducting an advanced search in the Web of Science (WoS), Thomson Reuters Core Collection. This search incorporated Social Sciences Citation Index (SSCI), Arts and Humanities Citation Index (A and HCI), Science Citation Index Expanded (SCI-EXPANDED), and Conference Proceedings Citation Index-Social Science & Humanities (CPCI-SSH). We chose WoS as the data source for two reasons. Firstly, WoS has established an independent and comprehensive editing process to ensure the excellent quality of the journals and has formed an unparalleled data structure based on more than 50 years of consistent, accurate, and complete indexing. The indexed journals in the Web of Science Core Collection have been carefully selected. Therefore, the articles indexed in WoS are of high quality. Secondly, WoS is CiteSpace’s primary data source. CiteSpace has been designed to work with WoS data. Datasets from other sources have to be transformed before they can be visualized in CiteSpace.

The following fields were used to retrieve the data:

TS = (metaphor*) AND (process* OR comprehen*)), which means that only articles with both “metaphor” and “process” or comprehen(sion) in the title or abstract, or keywords are retrieved.

Time span=1970–2022

Document Type=article OR review

(“*”is a wildcard in WoS that represents any group of characters, including no character. For example, metaphor*=metaphor, metaphors, and metaphorical, etc. In addition, the review articles in this research do not contain book reviews.)

Totally, 8358 papers were collected from 123 WoS categories, including experimental psychology, neurosciences, business, linguistics, management, music, nursing, and law. In our study, we specifically focused on metaphor-processing research in the fields of linguistics, psychology, and neurosciences. Therefore, we chose the WoS categories related to linguistics, psychology, neurosciences, literature, communication, sociology, philosophy, anthropology, religion, history, and law (i.e. “Linguistics” or “Language Linguistics” or “Psychology Experimental” or “Education Educational Research” or “Neurosciences” or “Psychology Multidisciplinary” or “Psychology Clinical” or “Psychology” or “Psychology Psychoanalysis” or “Psychology Educational” or “Psychology Applied” or “Psychology Social” or “Psychology Developmental” or “literature” or “communication” or “sociology” or “philosophy” or “anthropology” or “religion” or “history” and “law”). After excluding those works that were unrelated to metaphor processing, 3271 publications remained for further analysis.

Descriptive analysis

Before visualization by CiteSpace, we conducted a descriptive analysis of yearly publication trends. Our aim was to identify the most productive journals, authors, and institutions. These descriptive analyses were directly done on the data obtained from the WoS website. The number of works published each year has been given on the WoS website. We used SPSS software to obtain the annual trend of publications (see Fig. 1 ). The numbers of publications for each journal, author, and institution have also been given on the WoS website. We selected the top ten for analysis.

The diagram reveals the publication number for each year and the general trend.

CiteSpace analysis

The descriptive analysis of WoS provides only a basic overview of the research field. It cannot provide an exhaustive account of the research projects over previous decades and directions for future research. Previous reviews without bibliometric tools mainly relied on prior knowledge and subjective judgment. To address this problem, we used CiteSpace to examine the structures of the knowledge of metaphor processing that have been developed over the past years.

In this study, we used CiteSpace, a bibliometric analysis program developed by Chen ( 2004 , 2006 , 2017 ; also see Chen et al., 2010 ; Chen and Song, 2019 ). Bibliometric analysis offers an objective and quantitative method for examining published works in a certain area of research (Mou et al., 2019 , p. 221; Chen, 2020 ). CiteSpace is a Java application for analyzing co-citations and presenting them in the form of visual co-citation networks (Chen, 2004 ). CiteSpace is one of the most well-known bibliometric tools. It offers a variety of analyses, such as keyword analysis and reference analysis, to help academics identify current and upcoming research trends in a field (Mou et al., 2019 ). The bibliographic data files we collected from WoS were in the field-tagged Institute for Scientific Information Export Format. The “full record and cited references” was selected as the content. In this way, CiteSpace could easily identify the files. Once the files were loaded into the CiteSpace, the following procedural operations were performed on them: time slicing, thresholding, modeling, pruning, merging, and mapping (Chen, 2004 ).

In this study, we conducted two separate visualizing analyses of the data. One was a document co-citation analysis, which helped us to identify the important documents in metaphor processing research. A co-cited reference was called a node, and when several nodes were strongly related to one another, they formed a cluster. The other was a keyword co-occurrence analysis. The purpose of this analysis was to identify the most-discussed areas in research on metaphor processing.

Publication years, journals, productive authors, and institutions on metaphor processing

In the Web of Science core collection, the first article about metaphor processing we obtained was published in 1971 by Laurette ( 1971 ). There was no publication on metaphor processing in the years 1972, 1973, and 1974. From 1995 to 2022, more than 50 works were done each year. The maximum number of annual publications belongs to 2021 with 198 published works. Figure 1 presents the annual publications on metaphor processing. Overall, the results show a steady increase in publications on metaphor processing. Therefore, it can clearly be seen that metaphor processing has caught the attention of more and more researchers worldwide.

The 3271 articles or reviews that were examined in this study were published in a number of journals. Table 1 lists the 10 journals that published the highest number of papers in this area of research. With 116 publications on metaphor processing, Metaphor and Symbol , the only SSCI-indexed journal publishing works on metaphor research, was in first place among journals in terms of the number of publications. Frontiers in Psychology and Journal of Pragmatics were in second and third places, with 98 and 71 publications, respectively. The majority of the top 10 journals, as seen in Table 1 , are in the fields of psychology or neuroscience. When considering a submission, metaphor-processing researchers might use Table 1 to select appropriate journals for their papers.

The 10 authors having the highest number of publications in metaphor processing are listed in Table 2 . The author with the most papers published on metaphor processing was Mashal (36), followed by Faust (28) and Gibbs (26).

Table 3 lists the 10 institutions having the highest number of published works in metaphor processing. The University of California is at the top of this list with 131 publications in total, followed by the University of London with 76 articles and Bar Ilan University with 64 articles (Table 3 ).

Document co-citation analysis

A fundamental measure used by academic communities to assess the impact of a publication is the frequency of citations. The value of a published work and its impact on the field is at least partly dependent on the number of works that have been cited. We can identify the important documents in a knowledge domain by analyzing document co-citations. CiteSpace is an efficient tool that can conduct such analysis.

We analyzed document co-citations of 3271 publications collected from the WoS. We used CiteSpace to visualize the 3271 bibliographic recordings from 1970 to 2022. The top 50 papers having the highest number of citations in each 3-year were chosen using a time slice of three years. In order to include all the references cited in those documents regardless of when they were published, we set the Look Back Years (LBY) parameter to −1. Cutting off long-range citation linkages had a positive impact on the clarity of the results; it could increase the clarity of the network structure because long-distance links frequently go hand in hand with a spaghetti-like network. The results are shown in Fig. 2 . The cited publications and co-citation relationships across the entire data set were represented by 1055 distinct nodes and 5928 linkages, respectively. The top 10 articles in the area of metaphor processing research are shown in Table 4 .

The diagram of document co-citations reveals the top 10 most cited articles among the 3271 publications collected from the WoS.

Totally, between 1970 and 2022, 39 documents were cited more than 50 times. The top three most-cited publications in the world’s publications related to metaphor processing are classic books about Conceptual Metaphor Theory (CMT) in general, not about metaphor processing. The work that has received the most citations is “ Metaphors we Live by ” authored by Lakoff and Johnson ( 1980 ). This frequently cited book was a landmark that revolutionized research on metaphor processing. It contends that metaphor is a way of thinking, not just a rhetorical instrument. To put it simply, our conceptual system is fundamentally metaphorical. In contrast to earlier works that looked at metaphor as a purely linguistic figure of speech, this book emphasizes the conceptual nature of metaphor. It defines metaphor as a conceptual process in which a source domain is mapped into a target domain. For example, the conceptual metaphor ARGUMENT IS WAR, in which “argument” is the target and “war” is the source, can be used to explain a statement like “I defended my argument.” Since its introduction in 1980, Conceptual Metaphor Theory (CMT) has gained popularity across various disciplines. The second-most quoted work is also written by Lakoff and Johnson ( 1999 ). This book challenged the Western traditional philosophy by proposing Embodied Philosophy based on the premise that our actions and our languages are based on our bodily experiences. Embodied Philosophy contends that abstract concepts are largely metaphorical. Embodied Philosophy is thus considered as the philosophical basis of Cognitive Linguistics. The third most cited document is a monograph by Gibbs ( 1994 ). Gibbs illustrates that human cognition is inherently poetic and that figurative imagination is central to how we comprehend ourselves and our surroundings. It challenges the traditional understanding of the mind by demonstrating how figurative characteristics of language reflect the poetic structure of the mind. Psychology, linguistics, philosophy, anthropology, and literary theory ideas and research are utilized to demonstrate fundamental ties between the poetic structure of the mind and daily language use. This monograph discusses methods and findings of psycholinguistic and cognitive psychology research to assess current philosophical, linguistic, and literary theories of figurative language. CMT aroused the interest of scholars from different disciplines such as linguistics, cognitive science, neuroscience, and psychology. Scholars in neurolinguistics and psycholinguistics are particularly interested in the cognitive processing of metaphors.

The other publications in Table 4 are not about metaphor in general but about metaphor processing in particular. In an article entitled “An fMRI investigation of the neural correlates underlying the processing of novel metaphoric expressions”, Mashal et al. ( 2007 ) used functional magnetic resonance imaging (fMRI) to investigate the neural networks involved in the processing of related pairs of words that formed literal, novel, and conventional metaphorical expressions. Four different kinds of linguistic expressions were read by the participants, who then determined the relation between the two words (metaphoric, literal, or unrelated). The results showed that the degree of meaning salience of a linguistic expression, rather than literality or nonliterality, modulated the degree of left hemisphere (LH) and right hemisphere (RH) processing of metaphors. This supported the Graded Salience Hypothesis (GSH, Giora, 1997 , 2003 ), which predicts a selective RH involvement in the processing of novel and nonsalient meanings. In this study, the salient interpretations were represented by conventional metaphors and literal expressions, whereas the nonsalient interpretations were represented by novel metaphorical expressions. Right posterior superior temporal sulcus, right inferior frontal gyrus, and left middle frontal gyrus showed considerably stronger activity when the novel metaphors were directly compared to the conventional metaphors. These findings back up the GSH and point to a unique function of the RH in the processing of novel metaphors. Additionally, verbal creativity may be selectively influenced by the right PSTS.

In order to look into the neural substrates underlying the processing of three different sentence types, Stringarisa et al. ( 2007 ) combined a novel cognitive paradigm with event-related functional magnetic resonance imaging (ER-fMRI). Participants were required to read sentences that were either metaphorical, literal, or meaningless before deciding whether or not they made sense. The results of this experiment showed that various types of sentences were processed by various neural mechanisms. Both meaningless and metaphorical sentences activated the left inferior frontal gyrus (LIFG), but not literal sentences. Furthermore, despite the lack of difference between reaction times of literal and metaphoric sentences, the left thalamus is activated only in deriving meaning from metaphoric utterances. The authors attribute this to metaphoric interpretation’s flexibility and ad hoc concept formation. Their findings do not support the idea that the right hemisphere is primarily involved in metaphor comprehension, in contrast to earlier studies.

The two publications mentioned above used new research methods, such as fMRI and ER-fMRI. Additional research methods, such as repetitive transcranial magnetic stimulation (rTMS) by Pobric et al. ( 2008 ) and positron emission tomography (PET) by Bohrn et al. ( 2012 ), were also used in other highly cited papers.

Co-occurring terms analysis

Keywords of any paper present its theme and some kind of summary of the subject that is going to be discussed in it. The occurrence of two keywords in a piece of writing indicates that these words are closely related to one another in the content of the work. The prevailing view is that if two or more terms appear together more frequently, they are more closely related. Betweenness Centrality is one of the functions of CiteSpace that specifies the strength of the relation between two or more terms. This gives us the ability to predict the occurrence of a given term with other terms even in other related topics. If a keyword displays a high Betweenness Centrality value, the keyword may be very significant. In this study, the research areas and dominant topics can be determined utilizing keyword co-occurrence analysis.

We analyzed the keywords to identify the terms and phrases that had co-occurred in at least two separate publications. Highly-frequent terms can show hotspots in a specific field of research (Chen, 2004 ). In this study, we chose the slice length of 3 years, and we set the LBY to 5 years. The results showed language, comprehension, metaphor, figurative language and context were the top 5 key terms having the highest frequencies. The network of related keywords is shown in Fig. 3 , and the terms with a frequency of more than 40 are listed in Table 5 .

The keyword co-occurrence network diagram reveals the most popular keywords of metaphor processing research.

Cluster interpretations

We used CiteSpace to perform a cluster analysis on the basis of keyword co-occurrences. Totally, 528 nodes in the co-citation network with a 3-year time slice were obtained from the analysis. The seven greatest clusters in the research area of metaphor processing are displayed in Fig. 4 . Warmer colors represent more current research subjects, whereas cooler colors represent older research topics in the clusters.

The network diagram of the keyword co-occurrence cluster reveals the most significant clusters of metaphor processing research.

Table 6 shows the top 7 clusters of keywords in metaphor processing research. It is obtained by using index terms as labels for clusters. Also, the clusters were shown by log-likelihood ratio (LLR). The top 7 clusters are named students, figurative language, right hemisphere, embodied cognition, comprehension, N400 , and anger .

Cluster #0, as the largest cluster, is labeled as ‘students’. For native speakers, using and understanding metaphors is simple. However, understanding figurative statements might be challenging for non-native speakers. Littlemore et al. ( 2011 ) found that at a British university, second-language learners had trouble understanding 40% of metaphorical terms that were easily understood by native speakers. Results of another study showed that second-language learners tended to use metaphors incorrectly and in the wrong contexts (Kathpalia and Carmel, 2011 ). It may be challenging for second-language speakers to comprehend and generate metaphors since the metaphorical meaning of a term is developed in the social and cultural context of native speakers. Metaphorical expressions that are easily and automatically understandable for native speakers of a certain language may not be easily interpretable for second-language speakers of that language due to not having enough exposure to that language and culture (Kecskes, 2006 ). Therefore, one of the main concerns for second-language teachers is to enhance second-language learners’ ability in understanding metaphoric language and to use it efficiently in the cultural context of the second language. As a result, there is a lot of discussion in metaphor research about how to improve students’ ability in using metaphors. For instance, Hu et al. ( 2022 ) performed a randomized controlled trial to assess how metaphors affected the symptoms of anxiety in Chinese graduate students.

Cluster #1 is labeled as ‘figurative language’. This cluster shows that two key components of executive functions (working memory and inhibition) could play significant roles in figurative language processing. Since working memory holds information for a short period of time, it plays an active role in discourse comprehension. Therefore, this component of executive functions helps the individual use discourse clues and contextual information in the process of metaphor interpretation. Contextually relevant information and metaphorically relevant information are put together (Wilson and Sperber, 2012 ), enabling the individual to extract the intended metaphorical meaning from an expression. This is done by the active involvement of working memory. Also, the role of inhibition, as another component of executive functions, has been documented in many studies (e.g., Glucksberg et al., 2001 ). These two components can be in close interaction with one another in the process of metaphor comprehension.

Cluster #2 is labeled as ‘right hemisphere’ by LSI test (Chen et al., 2010 ). This cluster shows that functional magnetic resonance imaging has been a common technique in research on the role of the right hemisphere in metaphor processing. Over the past 20 years, researchers in the fields of neurolinguistics and psycholinguistics have intensively studied the role of hemispheres in metaphor processing. Some scholars have hypothesized that the right hemisphere (RH) may have a special role in the processing of metaphorical language. However, many behavioral studies (e.g., Bohrn et al., 2012 ; Faust and Mashal, 2007 ; Mashal et al., 2007 ; Mashal and Faust, 2008 ) have evidence suggesting that the processing of familiar or conventional metaphors requires more left-lateralized processing, compared to the processing of unfamiliar metaphors. Additionally, bilateral processing of traditional metaphors was also supported by the findings of several studies (e.g., Bambini et al., 2011 ; Diaz et al., 2011 ). These results lend credence to the Graded Salience Hypothesis (GSH), according to which semantic salience plays a key role in metaphor processing (Giora, 1997 , 2003 ). According to this hypothesis, conventional, frequent, recognizable, and prototypical meanings are simpler to process than less-prominent meanings. Therefore, the meaning of a conventional metaphor is more salient and more accessible than its literal counterpart. On the other hand, in a novel metaphor, the literal meaning is more evident and the figurative meaning is disclosed later with the support of contextual clues. The GSH claims that unlike novel metaphors, whose meanings are acquired through integration and inferential processes, conventional metaphors’ prominent meanings are stored in the mental lexicon. The GSH also predicts that the left hemisphere (LH) is more active in comprehending conventional and salient metaphorical meanings, while the right hemisphere (RH) is more active in comprehending innovative and non-salient metaphorical meanings (Giora, 2003 ).

Cluster #3 is labeled as ‘embodied cognition’. Theories of embodied cognition challenge the traditional theories of cognition that are based on amodal symbols. These theories offer new perspectives on human cognitive processes. These theories hold that simulation, situated action, and bodily states play a crucial role in cognitive processes. Cognitive linguistics gave rise to some of the first set of theories that supported grounded cognition. Theories of embodied language processing emphasized the role of body, situation, and simulation in language as opposed to the amodal theories of grammar that emerged in the Cognitive Revolution (e.g., Chomsky, 1957 ). The study of embodiment has caught the interest of researchers working in traditional cognitive science, who have started to incorporate the ideas of embodiment in their works. The role of embodiment in language processing was developed and promoted by George Lakoff, Mark Johnson, Mark Turner, and Rafael Núñez based on advancements in the field of cognitive science (Lakoff, 1987 ; Lakoff and Johnson, 1980 ; Lakoff and Johnson, 1999 ; Lakoff and Turner, 1989 ; Lakoff and Núñez, 2000 ). In their studies, they have found evidence suggesting that people draw on their knowledge of everyday physical phenomena to comprehend concepts. According to theories of embodiment and embodied language processing, cognition and cognitive processes are based on the knowledge that comes from the body. There has been an increase in interest in studying the relationship between embodied cognition and language over the last four decades. According to theories of embodied cognition, when people understand words, their sensorimotor systems are engaged in simulating the concepts the words refer to (Jirak et al., 2010 ). Lakoff and Johnson’s ( 1980 , 1999 ) conceptual metaphor theory (CMT) is one of the most prominent theories of embodied cognition. This theory holds that situated and embodied knowledge serves as the metaphorical foundation for abstract concepts. Specifically, Lakoff and Johnson ( 1980 ) argued that abstract concepts are metaphorically understood in terms of concrete concepts with the support of sensorimotor systems. Many studies in various languages have demonstrated how individuals frequently use physical metaphors to discuss abstract concepts. Literature also uses a lot of these metaphors (e.g., Turner, 1996 ). A crucial question is whether these metaphors only reflect linguistic convention or whether they genuinely represent how we think (e.g., Murphy, 1997 ). There is mounting evidence that these metaphors are essential to our thought (e.g., Boroditsky and Ramscar, 2002 ; Gibbs, 2006 ).

Cluster #4 is labeled as ‘comprehension’. One of the keywords in this cluster is the term context . This supports the key role of context in the process of metaphor comprehension. Context of a conversation can provide some information that contributes to metaphor processing (e.g., Steen et al., 2010 ). It helps the individual disregard non-relevant literal meanings and keeps the metaphorically relevant information to derive the intended metaphorical meaning.

Cluster #5 is labeled as ‘N400’. The N400 is a part of time-locked EEG signals called event-related potentials (ERP). It is a negative-going deflection that normally peaks over centro-parietal electrode sites and occurs 400 ms after the stimulus begins, though it can also last between 250 and 500 ms. The N400 is a typical brain response to words and other meaningful (or potentially meaningful) stimuli, such as visual and auditory words, sign language signs, images, faces, environmental sounds, and odors (Kutas and Federmeier, 2000 , 2011 ). During the past 4 decades, ERP has been one of the techniques most frequently employed in cognitive neuroscience research to examine the physiological correlates of sensory, perceptual, and cognitive activities associated with information processing (Handy, 2005 ). ERP is also widely employed in metaphor processing studies, along with other imaging techniques such as fMRI, PET, and MEG.

Cluster #6 is labeled as ‘anger’. This cluster includes the key terms figurative language and eye tracking . This suggests that the metaphoric conceptualization of some emotional states and emotional terms such as anger can be reflected in eye movements. Interestingly, some works have suggested that this can happen not only for emotion-related concepts but also for other categories of abstract concepts that are metaphorically described in terms of movement (e.g., Singh and Mishra, 2010 ).

This clustering of keywords offers an organized and clear picture of key concepts that have been involved in various lines of research on metaphor processing. This clustering shows which lines of investigation have had a strong relationship with one another in research on metaphor processing. Therefore, the suggested clustering of keywords in metaphor comprehension offers a map for research on metaphor processing. This can be a guiding tool for researchers to have a clearer idea and organized map of how various lines of research on metaphor processing intersect with one another.

Discussion and implication for future studies

Over the past 50 years, metaphor processing has been a widely discussed topic among scholars in various disciplines, particularly researchers in neurolinguistics and psycholinguistics. Through the aforementioned document co-citation analysis, co-occurring word analysis, and cluster visualization which were done by CiteSpace, this study showed that research on metaphor processing has mainly focused on hemispheric processing of metaphors, metaphor comprehension, the embodied cognition basis of metaphor processing, behavioral-experiments study, ERP method and other techniques (fMRI, PET, and MEG, etc.), and the comparision of metaphor processing of adults with children.

Results of this study showed that research projects on metaphor processing are mainly conducted by experiments, including behavioral experiments, ERPs, and other imaging techniques such as fMRI, PET, and MEG. However, there is some conflicting evidence in the research findings. For instance, many studies have shown no statistically significant difference between ASD and TD groups in the understanding of metaphors and figurative language (Hermann et al., 2013 ; Kasirer and Mashal, 2014 ; Mashal and Kasirer, 2011 ; Norbury, 2005 ). These results suggest that factors other than disease-specific traits may account for the differences in results between studies. In the past, it has been discovered that group matching strategy and general language proficiency can account for part of the between-study variability in figurative language comprehension (Kalandadze, et al., 2018 ). However, further pertinent variables need to be examined in order to fully explain the observed variabilities. One reason for these different results may be due to the different theories the researchers adhere to. Another reason for mixed results may be due to the task properties of those experiments.

As for the theories on metaphor processing, there are two models that are widely used to study the processing of metaphors, namely, the Direct Access View (Gibbs, 1984 , 1994 ; Gibbs and Gerrig, 1989 ) and the Graded Salience Hypothesis (Giora, 1997 , 2003 ). According to the Direct Access View, in metaphor processing, the non-literal meaning of the metaphor can be directly processed, without inferring and discarding the literal meaning in an initial stage. Based on the Direct Access View, the Parallel Hypothesis was proposed, which holds that understanding figurative language is not different from that of literal language. Therefore, it is not necessary to assume any special cognitive mechanisms to process figurative language such as metaphors (Glucksberg et al., 1982 ). However, the Parallel Hypothesis can only hold if the literal and figurative meanings are fully understood. When the literal and figurative meanings are inconsistent, the coexistence of literal and figurative meanings cannot be explained by the Parallel Hypothesis. This does not mean that the literal meaning is abandoned before it is processed. Rather, the context facilitates the understanding of the inconsistent literal meanings. Therefore, the Direct Access View also supports the Context-dependent Hypothesis, which holds that we have a direct understanding of the figurative meanings with the help of sufficient contextual information.

Another theory on metaphor processing that is widely used to support metaphor research findings is the GSH (Giora, 1997 , 2003 ). As mentioned, the GSH holds that metaphor processing is influenced by the degree of semantic prominence. That is, conventional, frequent, recognizable, and prototypical meanings are easier to assimilate than less-salient meanings. One prediction of GSH is that the right hemisphere (RH) is more active in perceiving creative and non-salient metaphorical meanings, while the left hemisphere (LH) is more active in comprehending conventional and salient metaphorical meanings (Giora, 2003 ).

While the Direct Access View holds that metaphorical meaning is directly accessible, the Graded Salience Hypothesis assumes that metaphorical meaning is activated after the activation of the salient literal meaning. Depending on which theoretical framework is taken for certain research, different and conflicting results may be obtained. However, it should be noted that metaphor processing is a complex phenomenon and a large number of factors may be involved in it. Therefore, a single theory may not be able to describe all aspects of metaphor processing for all types of metaphors. The Direct Access View can describe the processing of highly conventional metaphors and idiomatic expressions. In daily conversations, people can easily produce and understand conventional metaphors and idiomatic expressions automatically. But, in some cases, this theory fails to describe the processing of novel metaphors. On the other hand, the Graded Salience Hypothesis may provide a better picture of how novel metaphors are processed. Therefore, in order to explain the discrepancies in research findings, we may need to take broader frameworks. When a single theoretical framework cannot explain discrepancies, two complementary frameworks can be taken and combined to explain and reconcile the conflicting results. Furthermore, a given theoretical framework may be more applicable to certain groups of people. For example, the GSH may be more applicable to ASD than the AD group, while the Direct Access View may be more applicable to TD than the ASD group. In other words, types of metaphors (e.g., conventional vs. novel metaphor), features of comprehenders (ASD vs. TD group), and possibly many other factors determine which theory of metaphor processing is most applicable. Putting various theoretical frameworks together and trying to make broader theoretical frameworks is a potential solution for responding to some questions that have not been answered yet.

Another reason for the differences in results between studies on metaphor processing may be the task properties of those experiments. There is a consensus in the literature of behavioral and neuroimaging studies that factors such as clinical populations, task characteristics, response format (i.e., multiple-choice vs. verbal explanation task), and lack of linguistic context can affect participants’ capacity to interpret metaphors (Pouscoulous, 2011 , 2014 , Rossetti et al., 2018 ). For instance, when assessed with an act-out rather than a verbal explanation task, children with TD demonstrate earlier proficiency in metaphor understanding. This may be because verbal and other types of tasks place different demands on a child’s linguistic and cognitive abilities (Pouscoulous, 2011 ). A similar explanation for how people with ASD perform metaphor tasks is based on response format. For instance, people with ASD may grasp metaphors similarly to people with TD, but they may have more trouble conveying them orally because of problems with expressive language (Kwok et al., 2015 ). Other aspects of the metaphors may also play a role, such as the amount and type of contextual information that is available to interpret the expression, or the degree of familiarity with the expression (Pouscoulous, 2011 , 2014 ). By combining the preceding studies utilizing the techniques of systematic review and a meta-analysis, Kalandadze et al. ( 2019 ) collated the knowledge that is currently available concerning task properties. Their aim was to find out how task properties affect metaphor comprehension ability in people with ASD compared to people with TD. They discovered that previous studies had used various kinds of materials and tasks that were either created by the researchers who designed the studies or were adapted from earlier research. The possible impact of the task properties was rarely taken into account in the previous studies, despite the fact that the task properties varied widely. Degree of individual’s familiarity with the metaphor (conventionality/novelty), degree of complexity of syntactic structure, linguistic and non-linguistic context (physical context) of the metaphoric expression, modality of stimulus (e.g., audio, visual), response format (verbal or non-verbal), and timing of the task are important task properties that can affect results of studies and their interpretations. Therefore, in order to obtain more accurate results, these factors need to be taken into account.

Implication for future studies

Based on the discussion in the section “Discussion”, we suggest that two issues deserve more consideration in future studies on metaphor processing. The first one is the theories that are employed to support the findings of metaphor processing studies. As different theories on metaphor processing may generate different conclusions, it is suggested that researchers discuss the results from different theoretical perspectives, rather than a single theory.

The second issue that merits more consideration is task properties. Task properties are important but have been neglected. The existing research on metaphor processing has paid little attention to the relevance of task properties in performance on metaphor comprehension tasks. Therefore, we contend that task properties including response format and linguistic features (i.e., metaphor familiarity, the syntactic structure of the metaphor, linguistic context, and stimulus modality) should be carefully considered in future investigations on metaphor processing. The systematic review and meta-analysis by Kalandadze et al. ( 2019 ) revealed that some task properties, including metaphor familiarity, are more frequently taken into account than others when determining the impact of a task. The least studied property in previous research is syntactic structure. Also, research on metaphor processing has not done a good job to examine the influence of contextual information on different groups of people. In future metaphor processing studies, these task properties merit additional consideration. When creating and reporting task properties in metaphor studies, researchers need to be extremely careful.

It should be noted that metaphor processing is a complex and multidimensional process. Therefore, in order to obtain a clear picture of various aspects of metaphor processing, researchers of various fields need to collaborate in interdisciplinary research projects. Neuroimaging data collected by neurolinguistics experts, behavioral data collected by researchers in psycholinguistics and cognitive science, and even corpus-based data can be combined to offer a broader picture of metaphor processing. Various types of evidence can complement each other and fill the gaps. This is a crucial point that should be considered in future research on metaphor processing.

As noted by Han et al. ( 2022 ), metaphor processing has been the most studied research area in metaphor research. Since the 1970s, how metaphors are processed in the brain has been extensively investigated by scholars in linguistics, neurolinguistics, and psycholinguistics. However, up to now, bibliometric tools like CiteSpace have not been used to systematically review literature on metaphor processing. In our study, a total of 3271 bibliometric recordings were collected from the Web of Science Core Collection. These documents had been published between 1970 and 2022. The descriptive analysis revealed a yearly increase in the number of publications, indicating that metaphor processing has caught the interest of academics from a variety of disciplines. Metaphor and Symbol , the sole SSCI-indexed journal devoted to metaphor research, took the first position among journals in terms of publishing yield with 116 publications on metaphor processing. Mashal, Faust, and Gibbs are the three most prolific authors in terms of publications on metaphor processing.

These bibliometric analyses through the CiteSpace software showed that language, comprehension, metaphor, figurative language , and context were the five most frequent keywords. Also, the most prominent clusters were students, figurative language, right hemisphere, embodied cognition, comprehension, N400 , and anger . These findings showed that research on metaphor processing has largely focused on the hemispheric processing of metaphors, metaphor comprehension, and embodiment in metaphor processing. Behavioral experiments, ERP and other techniques, such as fMRI, PET, and MEG were the common techniques in metaphor processing research. The current review through CiteSpace indicates that putting various theoretical frameworks together and trying to make broader theoretical frameworks is a potential solution for responding to some questions that have not been answered yet. This review also suggests that in future studies on metaphor processing, task properties such as response format and linguistic features should be carefully taken into account.

Although the current study aimed to be comprehensive within its defined scope, it was subject to some inevitable limitations. Firstly, being limited to WoS documents was one of the limitations of this study. Other databases such as Scopus, Google Scholar, Index Medicus, and Microsoft Academic Search were not included in this study. Secondly, publishers’ labeling of document types was not always correct. Some articles presented as reviews by WoS, for example, were not review papers at all (Yeung, 2021 ). Thirdly, we used only one scientometric instrument. Fourthly, while several prospective papers have recently been published, these studies were not acknowledged. Furthermore, because of obliteration, the citation count for some earlier published works was low.

Nonetheless, this study comprises a ground-breaking bibliometric assessment of global research on metaphor processing and provides a clear overview of global publications related to metaphor processing. Hence, it can be a helpful source for researchers interested in metaphor and metaphor processing. The results of this review have both theoretical and practical implications for the study of metaphor processing and metaphor in general.

Data availability

All data analyzed during this study can be accessed at https://doi.org/10.7910/DVN/JFRP5W .

Bambini V, Gentili C, Ricciardi E, Bertinetto PM, Pietrini P (2011) Decomposing metaphor processing at the cognitive and neural level through functional magnetic resonance imaging. Brain Res Bull 86:203–216

Article   PubMed   Google Scholar  

Barsalou W (2008) Grounded cognition. Annu Rev Psychol 59:617–645

Bohrn IC, Altmann U, Jacobs AM (2012) Looking at the brains behind figurative language—a quantitative meta-analysis of neuroimaging studies on metaphor, idiom, and irony processing. Neuropsychologia 55:2669–2683

Article   Google Scholar  

Bowdle BF, Gentner D (2005) The career of metaphor. Psychol Rev 112(1):193–216. https://doi.org/10.1037/0033-295X.112.1.193

Boroditsky L, Ramscar M (2002) The roles of body and mind in abstract thought. Psychol Sci 13:185–188

Chen C (2004) Searching for intellectual turning points: progressive knowledge domain visualization. Proc Natl Acad Sci USA 101(Suppl 1):5303–5310

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Chen C (2006) CiteSpace II: detecting and visualizing emerging trends and transient patterns in scientific literature. J Am Soc Inf Sci Technol 57(3):359–377. https://doi.org/10.1002/asi.20317

Chen C (2017) Science mapping: a systematic review of the literature. J Data Inf Sci 2(2):1–40. https://doi.org/10.1515/jdis-2017-0006

Article   CAS   Google Scholar  

Chen C (2020) A glimpse of the first eight months of the COVID-19 literature on microsoft academic graph: themes, citation contexts, and uncertainties. Front Res Metr Anal 5:607286. https://doi.org/10.3389/frma.2020.607286

Article   PubMed   PubMed Central   Google Scholar  

Chen C, Ibekwe-SanJuan F, Hou J (2010) The structure and dynamics of cocitation clusters: a multiple-perspective cocitation analysis. J Am Soc Inf Sci Technol 61(7):1386–1409. https://doi.org/10.1002/asi.21309

Chen C, Song M (2019) Visualizing a field of research: a methodology of systematic scientometric reviews. PLoS ONE 14(10):e0223994. https://doi.org/10.1371/journal.pone.0223994

Article   CAS   PubMed   PubMed Central   Google Scholar  

Chomsky N (1957) Syntactic structures. Mouton, The Hague

Diaz MT, Barrett KT, Hogstrom LJ (2011) The influence of sentence novelty and figurativeness on brain activity. Neuropsychologia 49:320–330

Faust M, Mashal N (2007) The role of the right cerebral hemisphere in processing novel metaphoric expressions taken from poetry: a divided visual field study. Neuropsychologia 45:860–870

Article   CAS   PubMed   Google Scholar  

Gibbs RW (1984) Literal meaning and psychological theory. Cogn Sci 813:275–304

Gibbs RW (1994) The poetics of mind. Cambridge University Press, Cambridge

Google Scholar  

Gibbs RW (2006) Embodiment and cognitive science. Cambridge University Press, New York

Gibbs RW, Gerrig RG (1989) How context makes metaphor comprehension seem “special”. Metaphor Symbol Activity 4/3:144–158

Giora R (1997) Understanding figurative and literal language: The graded salience hypothesis. Cogn Linguist 8:183–206

Giora R (2003) On our mind: salience, context, and figurative language. Oxford University Press

Glucksberg S, Gildea P, Bookin HB (1982) On understanding nonliteral speech: can people ignore metaphors. J Verbal Learn Verbal Behav 21:85–98

Glucksberg S, Newsome MR, Goldvarg Y (2001) Inhibition of the literal: filtering metaphor-irrelevant information during metaphor comprehension. Metaphor Symbol 16(3-4):277–293. https://doi.org/10.1207/S15327868MS1603&4_8

Han Y, Peng Z, Chen H (2022) Bibliometric assessment of world scholars’ international publications related to conceptual metaphor. Front Psychol 13:1071121. https://doi.org/10.3389/fpsyg.2022.1071121

Handy TC (2005) Event related potentials: a methods handbook. Bradford/MIT Press, Cambridge, MA

Hermann I, Haser V, van Elst LT, Ebert D, Müller-Feldmeth D, Riedel A, Konieczny L (2013) Automatic metaphor processing in adults with Asperger syndrome: a metaphor interference effect task. Eur Arch Psychiatry Clin Neurosci 263(Suppl 2):S177–S187. https://doi.org/10.1007/s00406-013-0453-9

Holyoak KJ, Stamenković D (2018) Metaphor comprehension: a critical review of theories and evidence. Psychol Bull 144(6):641–671. https://doi.org/10.1037/bul0000145

Hu J, Zhang X, Li R, Zhang J, Zhang W (2022) A randomized controlled trial to evaluate the effect of metaphors on anxiety symptoms among Chinese graduate students: the mediation effect of worry. Appl Res Qual Life https://doi.org/10.1007/s11482-022-10107-2

Jirak D, Menz MM, Buccino G, Borghi AM, Binkofski F (2010) Grasping language—a short story on embodiment. Conscious Cogn 19(3):711–720. https://doi.org/10.1016/j.concog.2010.06.020

Kalandadze T, Bambini V, Næss K (2019) A systematic review and meta-analysis of studies on metaphor comprehension in individuals with autism spectrum disorder: do task properties matter? Appl Psycholinguist 40:1421–1454. https://doi.org/10.1017/S0142716419000328

Kalandadze T, Norbury C, Nærland T, Næss BK-A (2018) Figurative language comprehension in individuals with autism spectrum disorder: a meta-analytic review. Autism 22:99–117. https://doi.org/10.1177/1362361316668652

Kathpalia SS, Carmel HLH (2011) Metaphorical competence in ESL student writing. RELC J 42(3):273–290

Kasirer A, Mashal N (2014) Verbal creativity in autism: comprehension and generation of metaphoric language in high-functioning autism spectrum disorder and typical development. Front Hum Neurosci 8:615. https://doi.org/10.3389/fnhum.2014.00615

Kecskes I (2006) On my mind: thoughts about salience, context and figurative language from a second language perspective. Second Language Res 22:219–237

Kertész A, Rákosi C, Csatár P (2012) Data, problems, heuristics and results in cognitive metaphor research. Language Sci 34(6):715–727. https://doi.org/10.1016/j.langsci.2012.04.011

Kutas M, Federmeier KD (2000) Electrophysiology reveals semantic memory use in language comprehension. Trends Cogn Sci 4(12):463–470. https://doi.org/10.1016/s1364-6613(00)01560-6

Kutas M, Federmeier KD (2011) Thirty years and counting: finding meaning in the N400 component of the event-related brain potential (ERP). Annu Rev Psychol 62:621–647. https://doi.org/10.1146/annurev.psych.093008.131123

Kwok EYL, Brown HM, Smyth RE, Cardy JO (2015) Meta-analysis of receptive and expressive language skills in autism spectrum disorder. Res Autism Spectrum Disord 9:202–222. https://doi.org/10.1016/j.rasd.2014.10.008

Lakoff G, Johnson M (1980) Metaphors we live by. University of Chicago Press, Chicago

Lakoff G, Johnson M (1999) Philosophy in the flesh: The embodied mind and its challenge to Western thought. Basic Books

Lakoff G, Turner M (1989) More than cool reason: a field guide to poetic metaphor. University of Chicago Press, Chicago

Book   Google Scholar  

Lakoff G (1987) Women, fire, and dangerous things: what categories reveal about the mind. University of Chicago Press, Chicago

Lakoff G, Núñez RE (2000) Where mathematics comes from. Basic Books

Laurette P (1971) Metaphoric process and metonymic process-consideration of poem by valery. Linguistics 66:34–55

Littlemore J, Chen PT, Koester A, Barnden J (2011) Difficulties in metaphor comprehension faced by international students whose first language is not English. Appl Linguist 32:408–429

Mashal N, Faust M (2008) Right hemisphere sensitivity to novel metaphoric relations: application of the signal detection theory. Brain Language 104:103–112

Mashal N, Faust M, Hendler T, Jung-Beemane M (2007) An fMRI investigation of the neural correlates underlying the processing of novel metaphoric expressions. Brain Language 100(2):115–126

Mashal N, Kasirer A (2011) Thinking maps enhance metaphoric competence in children with autism and learning disabilities. Res Dev Disabil 32:2045–2054. https://doi.org/10.1016/j.ridd.2011.08.012

Morsanyi K, Stamenković D, Holyoak K (2020) Metaphor processing in autism: a systematic review and meta-analysis. Dev Rev 57. https://doi.org/10.1016/j.dr.2020.100925

Mou J, Cui Y, Kurcz K (2019) Bibliometric and visualized analysis of research on major e-commerce journals using CiteSpace. J Electron Commer Res 20(4):219–237

Murphy GL (1997) Reasons to doubt the present evidence for metaphoric representation. Cognition 62:99–108

Norbury CF (2005) The relationship between theory of mind and metaphor: evidence from children with language impairment and autistic spectrum disorder. Br J Dev Psychol 23:383–399. https://doi.org/10.1348/026151005X26732

Pobric G, Mashal N, Faust M, Lavidor M (2008) The role of the right cerebral hemisphere in processing novel metaphoric expressions: a transcranial magnetic stimulation study. J Cogn Neurosci 20(1):170–181. https://doi.org/10.1162/jocn.2008.20005

Pouscoulous N (2011) Metaphor: for adults only? Belg J Linguist 25:51–79. https://doi.org/10.1075/bjl.25.04pou

Pouscoulous N (2014) “The elevator’s buttocks” metaphorical abilities in children. In: Matthews D (ed.) Pragmatic development in first language acquisition. Benjamins, Amsterdam

Rai S, Chakraverty S (2020) A survey on computational metaphor processing. ACM Comput Entertain 53:2. https://doi.org/10.1145/3373265 . Article 24, 35 pp

Rapp A, Leube D, Erb M, Grodd W, Kircher T (2004) Neural correlates of metaphor processing. Cogn Brain Res 20(3):395–402

Rossetti I, Brambilla P, Papagno C (2018) Metaphor comprehension in schizophrenic patients. Front Psychol 9:670. https://doi.org/10.3389/fpsyg.2018.00670

Schmidt G, Seger C (2009) Neural correlates of metaphor processing: the roles of figurativeness, familiarity and difficulty. Brain Cogn 71(3):375–386. https://doi.org/10.1016/j.bandc.2009.06.001

Singh N, Mishra RK (2010) Simulating motion in figurative language comprehension. Open Neuroimaging J 4:46–52. https://doi.org/10.2174/1874440001004010046

Steen GJ, Dorst A, Herrmann JB, Kaal AA, Krennmayr T (2010) Metaphor in usage. Cogn Linguist 21(4):765–796. https://doi.org/10.1515/cogl.2010.024

Stringarisa A, Medford N, Giampietro V, Brammer M, Davida A (2007) Deriving meaning: distinct neural mechanisms for metaphoric, literal, and non-meaningful sentences. Brain Language 100(2):150–162

Turner M (1996) The literary mind. Oxford University Press, New York

Wilson D, Sperber D (2012) Meaning and relevance. Cambridge University Press, New York

Yeung AWK (2021) Is the influence of Freud declining in psychology and psychiatry? A bibliometric analysis. Front Psychol 12:631516. https://doi.org/10.3389/fpsyg.2021.631516

Zhao X, Zheng Y, Zhao X (2023) Global bibliometric analysis of conceptual metaphor research over the recent two decades. Front Psychol 14:1042121. https://doi.org/10.3389/fpsyg.2023.1042121

Download references

Acknowledgements

The authors extend their appreciation to the Humanities and Social Science Research Projects of the Chinese Ministry of Education [Grant Number: 19YJA740044].

Author information

Authors and affiliations.

Foreign Language Research Department, Beijing Foreign Studies University, No. 2, North Street of West Sanhuan Road, Haidian District, 100089, Beijing, P. R. China

Zhibin Peng

School of Foreign Languages, University of Electronic Science and Technology of China, No. 2006, Xiyuan Avenue, West Hi-Tech Zone, 611731, Chengdu, Sichuan, P. R. China

Omid Khatin-Zadeh

University of Religions and Denominations, Pasdaran, 37185-178, Qom, Iran

You can also search for this author in PubMed   Google Scholar

Contributions

ZP performed the data collection and analyses, and wrote the first version of the manuscript; ZP and OK-Z contributed to the revision of the manuscript.

Corresponding author

Correspondence to Zhibin Peng .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Ethical approval

This article does not contain any studies with human participants performed by any of the authors.

Informed consent

Additional information.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

3,271 bibliometric recordings from the web of science, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Peng, Z., Khatin-Zadeh, O. Research on metaphor processing during the past five decades: a bibliometric analysis. Humanit Soc Sci Commun 10 , 928 (2023). https://doi.org/10.1057/s41599-023-02465-5

Download citation

Received : 06 June 2023

Accepted : 27 November 2023

Published : 09 December 2023

DOI : https://doi.org/10.1057/s41599-023-02465-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Have a language expert improve your writing

Check your paper for plagiarism in 10 minutes, generate your apa citations for free.

• Knowledge Base
• What Is a Metaphor? | Definition & Examples

What Is a Metaphor? | Definition & Examples

Published on August 11, 2023 by Eoghan Ryan . Revised on November 6, 2023.

A metaphor is a figure of speech that implicitly compares two unrelated things, typically by stating that one thing is another (e.g., “that chef is a magician”).

Metaphors can be used to create vivid imagery, exaggerate a characteristic or action, or express a complex idea.

Metaphors are commonly used in literature, advertising, and everyday speech.

The exam was a piece of cake.

This town is a desert .

Table of contents

What is a metaphor, types of metaphor, metaphor vs. simile, metaphor vs. analogy, allegory vs. metaphor, worksheet: metaphor vs. simile, frequently asked questions.

A metaphor is a rhetorical device that makes a non-literal comparison between two unlike things. Metaphors are used to describe an object or action by stating (or implying) that it is something else (e.g., “knowledge is a butterfly”).

Metaphors typically have two parts:

• A tenor is the thing or idea that the metaphor describes (e.g., “knowledge”).
• A vehicle is the thing or idea used to describe the tenor (e.g., “a butterfly”).

Sophia was a loose cannon .

There are several different types of metaphor.

Direct metaphor

A direct metaphor compares two unrelated things by explicitly stating that one thing is another. Direct metaphors typically use a form of the verb “be” to connect two things.

Ami and Vera are two peas in a pod.

Implied metaphor

An implied metaphor compares two unlike things without explicitly naming one of them. Instead, a comparison is typically made using a non-literal verb. For example, the statement “the man erupted in anger” uses the verb “erupted” to compare a man to a volcano.

The captain barked orders at the soldiers. [i.e., the captain was like an angry dog]

Extended metaphor

An extended metaphor (also called a sustained metaphor) occurs when an initial comparison is developed or sustained over several lines or paragraphs (or stanzas, in the case of a poem).

Extended metaphors are commonly used in literature and advertising, but they’re rarely used in everyday speech.

And all the men and women merely players.

They have their exits and their entrances,

And one man in his time plays many parts,

Mixed metaphor

A mixed metaphor is a figure of speech that combines two or more metaphors, resulting in a confusing or nonsensical statement.

Mixed metaphors are usually accidental and are often perceived as unintentionally humorous. Mixing metaphors can confuse your readers and make your writing seem to lack coherence.

She’s a rising star, and with the right guidance, she’ll spread her wings.

Dead metaphor

A dead metaphor is a figure of speech that has become so familiar due to repeated use that people no longer recognize it as a metaphor. Instead, it’s understood as having a straightforward meaning.

The guest of honor sat at the head of the table .

Metaphors and similes are both rhetorical devices used for comparison. However, they have different functions:

• A metaphor makes an implicit comparison between two unlike things, usually by saying that one thing is another thing (e.g., “my body is a temple”).
• A simile makes an explicit comparison between two unlike things, typically using the words “like,” “as,” or “than” (e.g., “you’re as stubborn as a mule”).

The old man’s beard was as white as snow .

There are two main types of analogy:

• Identical relationship analogies indicate the logical relationship between two things (e.g., “‘Up’ is to ‘down’ as ‘on’ is to ‘off’”).
• Shared abstraction analogies compare two unlike things to illustrate a point.

Metaphors are sometimes confused with shared abstraction analogies, but they serve different purposes. While metaphors are primarily used to make a comparison (e.g., “John is a caveman”), shared abstraction analogies are used to make an argument or explain something.

Metaphors are sometimes confused with allegories, but they have different functions:

• A metaphor makes an implied comparison between two unlike things, typically by stating that one thing is another (e.g., “time is money”).
• An allegory illustrates abstract concepts, moral principles, or complex ideas through symbolic representation.

Allegories are typically longer than metaphors and usually take the form of a story.

You can test your knowledge of the difference between metaphors and similes with the worksheet below. Choose whether each sentence contains a metaphor or a simile.

• Practice questions
• Answers and explanations
• You sing like an angel.
• The boxer is as strong as an ox.
• Hannah is a warrior.
• Your eyes are deeper than the ocean.
• Most of the time, you’re an angel. But you’re like a demon when you’re tired.
• This sentence contains a simile because it makes a direct comparison using the word “like.”
• This sentence contains a simile because it makes a direct comparison using the word “as.”
• This sentence contains a metaphor because it makes an implicit comparison by saying that something is something else.
• This sentence contains a simile because it makes a direct comparison using the word “than.”
• This sentence contains both a metaphor (“you are an angel”) and a simile (“like a demon”).

An extended metaphor (also called a sustained metaphor ) is a metaphor that is developed over several lines or paragraphs.

The following is an example of an extended metaphor in William Shakespeare’s Romeo and Juliet :

“But soft, what light through yonder window breaks?

It is the East, and Juliet is the sun.

Arise, fair sun, and kill the envious moon,

Who is already sick and pale with grief

That thou, her maid, art far more fair than she.”

A metaphor is a figure of speech that makes a non-literal comparison between two unlike things (typically by saying that something is something else).

For example, the metaphor “you are a clown” is not literal but rather used to emphasize a specific, implied quality (in this case, “foolishness”).

Cite this Scribbr article

If you want to cite this source, you can copy and paste the citation or click the “Cite this Scribbr article” button to automatically add the citation to our free Citation Generator.

Ryan, E. (2023, November 06). What Is a Metaphor? | Definition & Examples. Scribbr. Retrieved April 17, 2024, from https://www.scribbr.com/rhetoric/metaphor/

Is this article helpful?

Eoghan Ryan

Other students also liked, how to avoid repetition and redundancy, parallel structure & parallelism | definition, use & examples, grawlix | definition, meaning, use & examples, unlimited academic ai-proofreading.

✔ Document error-free in 5minutes ✔ Unlimited document corrections ✔ Specialized in correcting academic texts

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Front Psychol

Global bibliometric analysis of conceptual metaphor research over the recent two decades

1 School of Foreign Languages, Jiangsu University of Science and Technology, Zhenjiang, China

Xincheng Zhao

2 Center for Applied English Studies, The University of Hong Kong, Hong Kong, Hong Kong SAR, China

Associated Data

The original contributions presented in this study are included in this article/supplementary material, further inquiries can be directed to the corresponding author.

Conceptual Metaphor has been a prevalent theme in the linguistic field for the recent twenty years. Numerous scholars worldwide have shown interest in it and published many academic papers from various stances on this topic. However, so far, there have been few rigorous scientific mapping investigations. With the help of bibliometric analysis tool, we selected 1,257 articles on Conceptual Metaphors published from 2002 to 2022, as collected in the Web of Sciences Core Collection database, from unique cognitive perspectives. The global annual scientific output of Conceptual Metaphor, including the cited articles, sources, keywords, and research trends, will be examined in this study. The most notable findings of this study are the following. First, there has been an upward trend in Conceptual Metaphor research over the last two decades. Second, the five most prominent research groups on Conceptual Metaphors are in Spain, the United States of America, China, Great Britain, and Russia. Third, future research on Conceptual Metaphors may focus on corpus linguistics, neurolinguistics, psychology, and critical discourse analysis. The interdisciplinary study may enhance the growth of Conceptual Metaphors.

Introduction

There is an enormous amount of literature in numerous research areas in the times of big data (e.g., Faust et al., 2018 ; Alnajem et al., 2021 ; Bashir et al., 2021 ; Abdul et al., 2022 ). Much research, however, is dispersed and difficult to compile in an orderly and visible manner. Therefore, finding specific literature quickly and accurately relevant to the research issue has always been challenging. For example, since Lakoff and Johnson proposed conceptual metaphors (CM) ( Lakoff and Johnson, 1980 ), academic papers on the growth of CM have undoubtedly increased over the past decades, and they have helped to advance numerous facets of CM study. However, keeping up with everything published instantaneously becomes more problematic. In terms of CM research, there are increasing studies on linked subjects, and various reviews have been conducted (e.g., Allahmoradi, 2018 ; Holyoak and Stamenkovic, 2018 ; Bundgaard, 2019 ; Gandolfo, 2019 ; Tohidian and Rahimian, 2019 ; Kövecses, 2020 ; Bearman et al., 2021 ; Jensen et al., 2021 ; Abdul et al., 2022 ). However, these studies mainly focused on qualitative analysis. So far, little research has been done on a general picture of CM research. Therefore, this study uses Bibliometrix metrology software, a statistical package called R (Biblioshiny), to visually analyze academic articles on CM over the recent two decades. The prevalent topics in different areas on the present research status, research themes, and future research directions in CM will provide references for scholars to research CM and predict its future direction.

This study is a valuable resource for academics and researchers working in the CM field. Beginners interested in CM can be offered the information required to start their research. Experienced CM researchers can familiarize themselves with the advances in the field and promote collaboration and networking between institutions and authors.

CM research (2002–2022): A bibliometric analysis

Bibliometric tool.

Bibliometrics is an open-source tool for quantitative research in scientometrics 1 . It is a unique tool using the R programming language for statistical computation and graphics following a logical bibliometric process. The bibliometric tool is controlled by Bibliometrix and its web-based graphical interface based on Web of Science (WoS), Scopus, and Dimensions data ( Aria and Cuccurullo, 2017 ). The interface is spontaneous and well-systematized, and the main menu is separated consistently with the Science Mapping Analysis system. The set menu performs analysis from eight categories: data sets, sources, authors, documents, clustering, intellectual structures, conceptual structures, and social structures in Biblioshiny. Several document formats can be transmitted, maps can be transferred to Html or Pajek, and tables can be saved as pdf, excel, or printed. The R bibliometrics package was used to analyze data, which gives more objective and dependable assessments than other methods. By providing a structured analysis of a large body of knowledge, bibliometrics becomes helpful when there is an excellent volume of new information, conceptual developments, and data to process.

Today, the bibliometric tool is increasingly utilized in numerous fields. An amount of research has been conducted ( Ellegaard and Wallin, 2015 ; Donthu et al., 2021 ; Efron et al., 2021 ), including tsunami research ( Chiu and Ho, 2007 ; Jain et al., 2021 ), the circular economy ( Geissdoerfer et al., 2017 ; Bui et al., 2020 ; Luis and Celma, 2020 ; Alnajem et al., 2021 ), green supply chain management ( Fahimnia et al., 2015 ; Amirbagheri et al., 2019 ), deep learning for healthcare applications ( Faust et al., 2018 ; Saheb et al., 2021 ; Zhu et al., 2021 ), environmental hypothesis ( Sarkodie and Strezov, 2019 ; Bashir et al., 2021 ; Hashemi et al., 2022 ), and COVID-19 research ( Verma and Gustafsson, 2020 ; Yan et al., 2021 ). Waltman et al. (2010) assert that scholars frequently mix mapping and clustering techniques when analyzing bibliometric networks. They employ bibliographic data from publishing databases to build structural pictures of scientific domains ( Zupic and Čater, 2015 ). The growing number of papers using bibliometric analysis across all disciplines suggests that it meets the desire of researchers who want proper research based on a wealth of literature.

Attributable to its reliable and scalable statistics, the bibliometric tool has compelling features and is becoming increasingly important in research. In contrast to other methods, it can introduce a systematic, transparent, and repetitious review procedure based on statistical assessments of science, or scientific activity. It adapts when the focus on empirical inputs results in extensive, dispersed, and contentious research streams, making it a particularly effective tool for science mapping ( Aria and Cuccurullo, 2017 ). This tool allows us to infer CM research trends and various themes researched, identify shifts in the boundaries of disciplines, find the most prolific scholars and institutions, and provide the large picture of prevailing research. Popular and thorough bibliometric analysis allow us to sift through and make sense of massive amounts of scientific data ( Donthu et al., 2021 , p. 285). This study aided us in analyzing the nuances of the CM research field’s evolutionary process and sheds light on its developing regions.

Conceptual metaphor research

In cognitive linguistics, conceptual metaphor (CM) refers to comprehending one thought or abstract concept in terms of another. Since 1980, theoretical clarification of CM has been the topic of extensive studies and lengthy introductions by many researchers (e.g., Lakoff, 1987 ; Lakoff and Turner, 1989 ). They showed great interest in CM research. According to Lakoff and Johnson, the mechanism of CM is as the following. “In a metaphor, there are two domains: the target domain and the source domain; in addition, metaphoric mapping is multiple. Two or more elements are mapped to two or more other elements. Image-schema structure is preserved in the mapping.” ( Lakoff and Johnson, 2003 , p. 266). They argue that metaphor is internally structured, and its meaning is derived from transferring specific characteristics from the original to the new field. As the surface manifestation of this mapping, the metaphorical expressions might be words, phrases, or whole sentences. The source domain is a somewhat tangible or, at the very least, strongly organized realm that often derives from our everyday experience. However, the target domain is where the metaphor is applicable and originates, a somewhat more abstract or unorganized field using unknown notions. Metaphor is the process of understanding one idea in a target domain via the other in a source domain. For instance, in the frequent metaphorical expression LOVE IS A JOURNEY ( Lakoff and Johnson, 2003 , p. 45), the target domain “LOVE” is abstract and difficult to construe. “JOURNEY” is the realm of the source. “LOVE” somewhat maps the “JOURNEY” structure in the following procedure: departure, on the way or lost way, and destination. As the conceptual metaphor is a mental construct, it is only meaningful when represented in more tangible elements. Therefore, this sentence consists of many metaphorical analogies that form a unified inner structure. “LOVE” and “JOURNEY” are strongly connected in this sense. The idea of “love as a journey” shapes the conceptualization of love itself. Even though “love” might be understood in ways other than a journey, we use this comparison to impact our understanding and attitude toward LOVE. This approach is how we may see, experience, participate in, and refer to LOVE IS A JOURNEY. Metaphor infuses our language, daily lives, and actions. Because the mind is experienced, our cognition is experiential. Remarkably, human cognition derives from our personal experience of the external world, shaping our perspective on the outer world.

Undeniably, the growth of academic papers on CM has contributed to the advancement of CM research. Numerous scholars have conducted a significant amount of research into CM from various perspectives, such as psycholinguistic metaphor research ( Murphy, 1996 ; Gibbs, 2013 ; Qiu et al., 2022 ); deliberate metaphors and embodied simulation research ( Gibbs, 2006 ; Cuccio and Steen, 2019 ; Cuccio et al., 2022 ); conceptual conflicts in metaphors and translation ( Prandi, 2017 ; Rizzato, 2021 , 2022 ); corpus-based metaphor research ( Sinclair, 1991 ; Charteris-Black, 2000 , 2004 ; Semino, 2002 ; Deignan and Potter, 2004 ; Allen, 2006 ; Fabiszak, 2007 ; Tissari, 2010 ; Shutova et al., 2013 ; Burgers and Ahrens, 2018 ; Zhao and Zhou, 2019 ; Zhao et al., 2019 , 2020 ; Silvestre-López, 2020 ; Bosman and Taljard, 2021 ; Kazemian and Hatamzadeh, 2022 ), critical metaphors in discourse analysis ( Charteris-Black, 2004 ; Ferrari, 2007 ; Musolff, 2012 ) and metaphors in classroom teaching ( Thomas and McRobbie, 2001 ; Andreou and Galantomos, 2008 ). There is also increasing research on various reviews of CM research (e.g., Allahmoradi, 2018 ; Holyoak and Stamenkovic, 2018 ; Bundgaard, 2019 ; Gandolfo, 2019 ; Tohidian and Rahimian, 2019 ; Kövecses, 2020 ; Bearman et al., 2021 ; Jensen et al., 2021 ; Abdul et al., 2022 ).

However, researchers need help to pinpoint the research status and anticipate research trends rapidly and correctly. Keeping up with articles published instantly also becomes increasingly challenging. By using a bibliometric analysis, this knowledge map will be an invaluable resource for beginning researchers to learn more about information and study results to start their investigation as soon as possible. Additionally, this study will identify future research gaps and find potential cooperators for seasoned scholars. In addition, this study will provide some rating agencies with a trustworthy benchmark to assess the effectiveness of authors, institutions’ sources, and nations in CM research. Nevertheless, there has not been a thorough visual of CM studies so far. The bibliometric analysis of Bib text provides extra data statistics, including author, affiliation, and keyword ( Fahimnia et al., 2015 ). Therefore, this study will fill the gap by analyzing the state of CM’s research over the past 20 years, its current focus areas, and future research directions.

Methodology

Research questions.

With the bibliometric tool, this study aims to provide an overall picture of CM research over the recent two decades and address the following three questions:

• (1) What was the basic information about the development of international CM research in the past two decades?
• (2) What is the present situation, including yearly scientific advancements, subject orientations, most renowned authors, and the most pressing issues in CM research?
• (3) What predictions may be made regarding its future development based on a bibliometric study?

Data source

All the data in this study were obtained from WoS Core Collection. It is the platform’s flagship resource, covering over 21,000 peer-reviewed, high-grade scientific articles (containing Open Access journals), more than 205,000 conference proceedings, and more than 104,000 editorially selected book 2 .

It offers more reliable journal coverage of scholarly published articles ( Birkle et al., 2020 ) than any other databases like Scopus and Google.

Data collection

This study discerningly chose the WoS that confined the data from 2002 to 2022. The literature data gained comprised the whole archives, such as the author’s name, source year, abstract, keywords, citation frequency, DOI number, and references in the article. Data collection consisted of three stages. The first was data reclamation. We prudently chose the papers and early access collected in the arts and humanities citation index (AHCI) and the SSCI to evaluate research questioning. We scrutinized the principal articles consistent with the research topic. The second step was data scrubbing. We sifted papers discreetly to avert data duplication. In the third step, documents were downloaded and compacted. We downloaded 1,000 files the first time and 257 the second time. Subsequently, the two files were compacted using bibliometric instruments. Currently, diverse instruments are accessible to present visual studies, such as CitNetExplorer, CiteSpace, and VOSviewer. This study selected a Biblioshiny program to obtain an overall visual picture of CM study in the past two decades because it has unique features. The set menu in Biblioshiny presents analysis from source, author, and document dimensions. Additionally, this menu offers conceptual, intellectual, and social knowledge structures. Maps can be exported to HTML or Pajek, tables can be copied to the clipboard or saved as Excel or PDF files, and maps can be printed. We analyzed the data using the Rstudio software and the bibliometric R-package version 4.2.0. The bibliometric analysis was first enabled in the R environment using the following command code:

• install. Packages (“bibliometrix,” dependencies = TRUE)
• library (bibliometrix)
• biblioshiny()

The Biblioshiny web interface was presented once the Google Chrome browser started with the above code. Raw WoS data were imported and analyzed using the Biblioshiny. We then went on to describe and evaluate the critical results of the study, which were shown with the statistics and pictures. This study employed pertinent authors, institutions, countries, articles, top highly cited publications, keyword co-occurrence, word clouds, thematic maps, trend topics, and conceptual framework to answer the above three research questions of the study.

Results and discussion

Position of cm research in the past two decades.

Table 1 presents the key findings of the entire CM research from January 1, 2002, to July 10, 2022. In total, 1,257 documents were present. There were 317 sources for the CM research, including books, journals, and other materials. The average number, like years from publication, citations per document, and citations per year per document, were 7.49, 8.228, and 0.7505. The number of references cited in the studies reached 33,265, demonstrating the popularity of CM research over the previous two decades. The 956 papers represented the most significant categories of published documents. The author’s keywords and the plus were 3,130 and 846, respectively, in terms of the document contents. It demonstrates the variety of topics covered by CM research and 1,544 contributors to CM studies from 2002 to 2022. There were 613 authors of single-authored documents and 931 authors of multi-authored documents. The single-authored documents are 776, showing that the scholars are highly interested in this area. The documents per author were 0.814, while authors per document, co-authors per document, and collaboration index were 1.23, 1.59, and 1.94. It indicates that more scholars concentrate on CM research, and the direction conducted by multiple authors was the most important means for CM research in the past two decades (see Table 1 ).

Main information about data.

Annual scientific production

A highly intriguing phenomenon has been discovered in annual scientific production. Figure 1 depicts the dynamics of document creation. The number of papers published annually was balanced from 2004 to 2006 and steadily rose from 2002 to 2022. The most productive year for the output was 2020, with 119 publications, including Gender, Ideology, and Conceptual Metaphors: Women and the Source Domain of the Hunt ( Maestre, 2020 ) and Conceptual Metaphors Leading to Some Names of Anger in the Indo-European Languages (With Focus on the Romance Languages) ( Georgescu, 2020 ). Notably, this number has steadily increased, with a yearly growth rate of 2.05 percent. The number of studies on CM and the total number of articles published has expanded significantly over the past 5 years. The yearly variation in literature production may represent the shift in the research subject, research interest, depth, and future development direction. CM has been a prevalent topic in the linguistic field over the past two decades, and it may continue to be a future topic in this field. In other words, the CM in Cognitive Linguistics has garnered great academic interest over the past two decades.

Annual scientific production.

Analysis of cited documents

Average annual citations.

In Figure 2 , we can see the typical annual number of article citations. The most significant number of citations was 2,796 in 2006, while the least was 0.546 in 2019. Typically, the yearly average citation rate of recent articles is low. There is a surprising phenomenon: the citation rate of CM articles in 2006 reached a peak, but the publications in 2006 were low. Therefore, the citation rate may be more relevant to the articles’ quality and themes rather than their quantity. The top four average annual citations articles in 2006 are the following, Metaphor Interpretation as Embodied Simulation ( Gibbs, 2006 ), The Emergence of Metaphor in Discourse ( Cameron and Deignan, 2006 ), Does Understanding Negation Entail Affirmation: An Examination of Negated Metaphors ( Hasson and Glucksberg, 2006 ), Metaphoric competence, Second Language Learning, and Communicative Language Ability ( Littlemore and Low, 2006 ). The constant average citation per year after 2014 in Figure 2 shows that CM research has lately had a stable development.

Average article citations per year.

Most global citated articles

Figure 3 shows CM’s top 20 most globally cited documents from 2002 to 2022. According to Figure 3 , Gibbs’s article Metaphor Interpretation as Embodied Simulation ( Gibbs, 2006 ) was most passionately cited with 198 citations, hierarchical first among all other documents. In this study, Gibbs (2006) claims that part of our ability to make sense of metaphorical language, both individual utterances and extended narratives, resides in the automatic construction of a simulation whereby we imagine performing the bodily actions referred to in the language. As Time Goes by: Evidence for Two Systems in Processing Space → Time Metaphors ( Gentner et al., 2002 ) has 189 total citations and demonstrates that individuals employ spatial metaphors in temporal thinking. The metaphoric systems’ status implications are examined in it. With 169 citations, Gibbs et al. (2004) review the empirical evidence and discuss the methodological strategies employed by linguists and psychologists seeking connections between embodiment and CM. Subsequently, The Emergence of Metaphor in Discourse ( Cameron and Deignan, 2006 ), Literal vs. figurative language: Different or Equal ? ( Giora, 2002 ). Other significant subjects of CM research are the relationship between CM and metonymy, studying psychology and politics of metaphors, and CM based on language theory. Most of the literature that generates the most citations has been published for more than ten years, indicating that the topic and authority of the publication may be the reason for the number of citations. Figure 3 also shows that the years of highly cited literature on CM were 2006, 2002, and 2004, representing that CM has made a breakthrough in development during these years. Besides, Figure 3 suggests that a longitudinal study of how CM works over time is crucial to scrutiny. In general, the more citations an article has, the more influential it will be in the CM field. Moreover, Figure 3 proves that Gibbs’ article published in 2006 was the most relevant document contributing to the CM research. The research of CM is closely related to human psychology and cognition, and it may be more concerning and exciting to scholars when they conduct empirical research.

Most global cited documents.

Source growth

Figure 4 depicts the source dynamics of the top five journals from 2002 to 2022. Regarding the number of articles, Figure 4 shows a significant increase trend, with the peak in 2022 and the lowest in 2002. The corresponding maxima are the following: Review of Cognitive Linguistics, Metaphor and Symbol, Cognitive Linguistics, Journal of Pragmatics , and Journal of Literary Semantics . The increase in sources illustrates the main application areas of CM research over the past two decades and its multidisciplinary development trend. As indicated in Figure 4 , Journal of Review of Cognitive Linguistics has published CM articles in recent years with the highest growth rate, particularly between 2012 and 2022. This Journal’s quick expansion indicates that several experts enthusiastically pursue the debate and research on CM. Despite being among the top five, as shown in Figure 4 , Journal of Literary Semantics had a steadily increasing number of CM papers published from 2002 to 2022. Only 21 articles were published in this journal in 2022, but there were 110 articles in Reviews of Cognitive Linguistics . The number indicated that the Journal’s discussion subject might diverge from the study category of CM. From 2002 to 2022, we judged from the growth trend of article sources that CM’s research showed a sound momentum of rapid progress over the last two decades.

Source dynamics.

Authors, affiliations and countries

Prolific authors.

Gibbs was the most significant researcher, who published 17 articles and ranked first in document number on CM, concentrating on the embodied metaphor and mapping in cognitive linguistics in terms of the author’s output from 2004 to 2022. Gibbs’s articles On the Psycholinguistics of Sarcasm and How to kick the Bucket and not Decompose: Analyzability and Idiom Processing , with more than 200 references to Spilling the Beans on Understanding and Memory for Idioms in Conversation . Gibbs is committed to studying embodied metaphors and mapping in cognitive linguistics and makes significant contributions to the CM research. Following Gibbs, Yu published nine documents mainly scrutinizing the spatial subsystem of moral metaphors in English. De Mendoza Ibanez represents Spanish research on CM with nine articles. He explores metaphors concerning cognitive prominence and conceptual interaction issues.

Moreover, he deals with the problems of constraints on metaphor and proposes three complementary kinds of constraints. Over the past two decades, these three authors were the most productive and essential in the CM research field. They are vital scholars, and their views may provide a theoretical and practical framework for further research.

Most relevant affiliations and countries

Most relevant affiliations can present the top most relevant affiliations according to the number of articles about CM. The University of La Rioja, the University of California, Santa Cruz, the University of Birmingham, Castile La Mancha University, and Guangdong University of Foreign Studies were the five most relevant affiliations by producing 68, 30, 28, 28, and 25 articles in the past two decades, respectively. They are also the bases for linguistic research. The result derived from the cooperative efforts of various institutions and was focused on CM subjects.

Figure 5 shows the countries of the top 20 corresponding authors. The collaboration of authors of the same nationality was far more than that between nations. According to Figure 5 , the top 20 countries contributed a lot to the research of CM during the past two decades. Among them, Spain contributed the most to it, with 171 publications, followed by the United States (138), China (131), the United Kingdom (87), and Russia (83). These countries occupied the top five in the WoS Core collection. The result showed that CM research attracted the attention of researchers in these countries in the past two decades. To some extent, the publications of these authors will benefit CM’s future development.

Corresponding author’s country. Multiple country publications (MCP), the number of papers co-authored with authors from other countries; SCP, the number of papers co-authored by authors of the same nationality. The MCP ratio represents the ratio of international cooperation.

Table 2 lists countries, average article citations, and total citations for the top 20 relevant nations. Table 2 shows that the United States made the most considerable contribution to CM research, with 3,203 total citations and an average article citation rate of 23.210. The United Kingdom and Spain came next with 1,378 and 1,112 total citations, respectively. It implies that scholars from the top three nations show great interest in CM research. A particular topic of study was directly tied to the context of the nation. Therefore, the United States, the United Kingdom, and Spain contribute to CM research development and associated linguistic issues with more attention. The fact that more nations, including the Netherlands, China, and Italy, are paying attention to CM research shows how prevalent it has become over the past two decades.

Most cited countries.

We can also see from Table 2 that although the corresponding authors of CM research in the United States ranked second, their citation rate ranked first. Similarly, the corresponding authors in the UK rank fourth, while the citation rate of their authors ranks second. Therefore, the number of correspondents does not have a one-to-one proportional correspondence with their citation rate. From this, we can infer that the citation rate may relate to the article’s quality and themes.

Conceptual structure

Figure 6 presents the current status of thematic groups in CM research. Thematic maps illustrate a particular topic and help reveal geospatial patterns and relations ( Schaab et al., 2022 ). A thematic map is separated into four quadrants grounded on the degree and density of centrality. High density and centrality in the upper right quadrant represent well-developed Motor Themes in the CM research area. In this quadrant, many themes comprised the emphasis and center of CM research, such as “comprehension,” “conceptual integration,” “words,” “metaphors,” “English,” “space,” and “language.” These themes had outstanding growth in the past two decades. The second quadrant’s high density and low centrality imply niche themes with good development prospects but a limited influence on the research field. Although scholars have created a “mechanisms” research group, its prospects are unsure. Subject clusters have poor centrality and density in the lower-left quadrant. It implies that different types of “semantics,” “vocabulary,” “metaphor,” “discourse,” “conceptual metaphor,” and “metonymy” are marginalized. It suggests they are new or waning themes. The fourth quadrant’s high centrality and low density indicate that “mind,” “children,” “deficits,” “idioms,” and “memory” are the primary topics in CM. Their theoretical systems are more thorough and mature, and these core topics may provide the theoretical foundation, reasoning, and technique for CM research.

Thematic map.

Figures 7 , ​ ,8 8 graphically depict the development of the CM study subjects. 2016 was the dividing line and the two time periods were 2002–2016 and 2017–2022. The topics from 2002 to 2016 may be summed up as “context,” “conceptual integration,” “language,” and “time,” with researchers focusing on conceptual integration. According to the CM hypothesis, metaphor incorporates two cognitive domains, while abstract blending theory theoretically converts two cognitive parts into four mental spaces ( Fauconnier and Turner, 1998 ). It may more precisely characterize people’s psychological processes while using metaphor. In Conceptual Blending Theory, the creation and functioning of conceptual blending are creative. This theory may thus explain not just established mental metaphors but also novel metaphors. Individuals’ daily communication and understanding process is an innovative online mapping and integration process. The relationship between online mapping and fixed mapping is tight. The idea of conceptual blending is comprised of four cognitive domains. According to the theory, the human thinking mode is not a direct, unidirectional, and absolute mapping of the source domain to the target domain but rather a dynamic integration process in which the shared mental schema is a generic space. The two input spaces of the source domain and target domain are bidirectionally mapped to the blending space. Mental space, not the cognitive part, is the fundamental unit of cognitive structure in conceptual blending. Mental space is an abstract area created when individuals think, act, and communicate, intending to achieve local comprehension and action. It is only a transient framework comprised of conceptual aspects like time, belief, desire, possibility, virtuality, place, and reality and depends on the cognitive field, a broader and more fixed knowledge structure associated with a particular cognitive area. It reflects the specific mental schema generated by the cognitive domain and it is dynamic, adaptable, and active throughout the thought process.

Time slice 1: thematic evolution during 2002–2016.

Time slice 2: thematic evolution during 2017–2022.

From 2017 to 2022, the nature of the mental processes involved in metaphor comprehension was the focus of debate ( Stamenković et al., 2019 ), with dispute focusing on the relative function of common analogical reasoning versus language-specific conceptual blending. The accompanying research indicated that the blending theory framework had explanatory power and practical use.

Figure 9 shows that the three fields plot can comprehensively analyze the relationship between measurement indicators of different literature and build a comprehensive network map. According to the statistics, among the periodicals published from 2012 to 2022, Metaphors We Live By ( Lakoff and Johnson, 2008 ) was cited first, followed by Women, Fire and Dangerous Things : What Categories Reveal about the Mind ( Lakoff, 1987 ). The middle part of the Three-fields Plot is the Citation Source. Cognitive Linguistics ranks first in this field, followed by Metaphor and Symbol and Metaphors We Live By . Cognitive Linguistics is the first citation source, and its corresponding citations are mainly George Lakoff’s books, which shows the authority of George Lakoff, the founder of CM theory, in this field. On the right is the authors’ keyword part. We can see that “metaphor” ranks first on the pyramid, and “metonymy” ranks second. which is consistent with the following Co-occurrence Network (see Figure 11 ).

Three-field plot. The middle field is cited sources, the left refers to references, and the right refers to the author’s keywords.

Conceptual structure map. The map is split in half. Clusters are indicated by color, proximity between keywords mean their relationship, the vertex is an illustration of the term in question, and the node’s size is proportional to the frequency with which it appears.

Research topics in CM

Research in the recent two decades.

Content analysis was employed to illustrate the CM research issues. Word Cloud, a thematic map of word growth, a conceptual structure map, and the co-occurrence of the author’s keywords were utilized to show study subjects in CM research during the recent two decades.

Word clouds are a valuable tool for providing overviews of texts and visualizing relevant words ( Herold et al., 2019 ). Word Cloud was based on the author’s keywords for CM research between 2002 and 2022. With a visual depiction of the Biblioshiny, words with greater volume and keyword density were shown in a bigger and clearer typeface. Word Cloud was used to evaluate commonly used terms in CM research to reveal study subjects. To be more precise, the frequency of usage of a term increased according to its centrality. Based on the author’s keywords, we selected the top 20. First, Figure 10 shows that “metaphor” was the most frequently used term in the authors’ publications, with 143 times in the extracted database, followed by “language (129),” and “comprehension (61).” The number indicates that CM was a vital study issue in cognitive science over the last two decades. The other terms “discourse,” “mind,” and “metonymy” were also used extensively as keywords by writers. It demonstrates that these were essential subjects in the CM field.

Word cloud. Different words were colored differently, and the size and placement of the phrase denoted their frequency. The size of the colored words depicts the frequency of their occurrence.

Conceptual structure map

Researchers may utilize Biblioshiny for Bibliometrix’s Conceptual Structure Map for multiple correspondence analysis (MCA), which aids in sketching a conceptual structure of the area and locating groupings of texts that express similar concepts. Using MCA, one may do a mathematical and graphical analysis of seemingly multivariate data ( Greenacre and Blasius, 2006 ). Figure 11 displays the results of MCA’s clustering on the keywords. The terms “metaphor,” “conception,” and “conceptual integration” often occur in the red set. “Frames,” “future,” and “construal” were added to the blue grouping. Figure 11 shows progress has been made toward developing a significant study subject of CM, and specific research issues around CM have been advanced a fair amount. Metaphors in “time,” “space,” “context,” “emotion,” “anger,” and “expressions” have been analyzed. Multiple fields, such as “pragmatics,” “semantics,” and “memory,” have been thoroughly researched in terms of CM. In addition, the CM research was related to “cognition,” among other things, and not only “discourse,” “metonymy,” “corpus,” and “mind.” Many scholars have conducted studies on metaphor from the perspective of psycholinguistics. Cuccio and Steen (2019) emphasize that attention is a crucial notion in defining deliberateness in metaphor processing because it is the attention we pay to the source domain of a metaphor in working memory that makes a metaphor a deliberately processed metaphor. Gibbs (2013) describes a few complications in psycholinguistic investigations of metaphor and explains the variability of study results. It is common knowledge that engaging in insightful metaphor analysis can be helpful in better comprehending how psychological trauma is conceived. As [ Cuccio et al. (2022) , p. 1] go, “we need to explain how we use symbols and how we make meanings out of them.” Increasingly, scholars talk about the construal of CM, such as the role of context in the interpretation of CM ( Zhao, 2008 ; Zhao et al., 2020 ).

Co-occurrence of keywords plus

Figure 12 presents four distinct clusters: blue, green, red, and purple. The blue cluster focuses on “language,” “representation,” and “communication;” the green cluster emphasizes “comprehension,” “mind,” “idioms,” and “words,” and the red cluster emphasizes “metaphor” and “discourse.” Clustering in purple mostly depends on “time,” “space,” and “perspective.” Consequently, Conceptual Metaphor research emphasizes linguistic theory study, corpus empirical research, and discourse analysis. Critical Metaphor Analysis, also known as CMA, is a method that is typically applied to the process of analyzing metaphors in various critical discourses to reveal the feelings, attitudes, and thoughts that lie behind metaphors. Charteris-Black (2004) proposed “Critical Metaphor Analysis,” which combined pragmatics, cognitive linguistics, and critical discourse analysis. He argued that while cognitive semantics provided a suitable description of how humans comprehended metaphors, the social effect of ideology, culture, and history might give a more persuasive explanation for why specific metaphors were selected in contexts. “Discursive-pragmatic factors, as well as sociolinguistic variation, have to be taken into account to make cognitive analyses more empirically and socially relevant” ( Musolff, 2012 , p. 301). When it comes to addressing persuasion in text, CM, as it relates to emotion, is a crucial tool because it helps identify the ideological root and persuasive strategy of a given discourse ( Ferrari, 2007 ).

Co-occurrence of keywords plus. The four distinct clusters are blue, green, red, and purple.

Co-occurrence network development relies heavily on correlation inference. The co-occurrence network has several study and application disciplines, and each color refers to a field. Language is a cognitive tool and a product of human intellect. With the rise of multidisciplinary study, cognitive explanations for grammar creation, semantics, discourse, and metaphor have become widespread, founded on empiricism and cognitive science research. It attempts to explain that language phenomena conform to the human understanding of the brain and thinking, i.e., human language is the product of the human brain, and its construction principle is identical to that of other cognitive domains. Therefore, Figure 12 shows the most significant community, the blue “language.” The closer to the central district, the closer to the “Conceptual Metaphor.” The figure presents that the closest to “language” takes “metaphor” and “comprehension” as the keywords.

The co-occurrence of keywords analysis is a valuable method for constructing a comprehensive framework for comprehending the significant areas of CM study during the past two decades. Figure 12 illustrates a network of co-occurrence between keywords in different types of publications that were established. When two or more of an author’s keywords appeared together, it might indicate how often those terms appeared together in the same publication. Each period was represented as a node, and the greater the node’s size, the more times that the keyword was cited. The greater the thickness of the line that connected two nodes, the more often those terms appeared together.

In the same way, Figure 12 also displays five distinct groups, each representing a different hue. In particular, the terms “discourse” and “metaphor” often co-occurred and were distributed heavily in the red cluster. This suggests that “discourse,” “expressions,” and “English” were prioritized in the CM study, and CM research was practically inseparable from “mind” research. In the center of the purple circle stood the word “time,” but it was disconnected from the surrounding words. However, the connections to “space” and “perspective” were weak. In addition, this analysis discovered that “comprehension” and “metaphor” were often investigated together and that “words” and “idioms” research were linked based on the frequency with which these terms occurred in green nodes. The small size of the nodes and the scarcity of connecting lines suggested that these concerns had not been well-explored.

In conclusion, several terms were used in the investigation of CM. The terms “metaphor,” “language,” “comprehension,” and “English” featured prominently and were among the most often co-occurring in the text. This demonstrated that these issues were central to the CM study. Words like “brain,” “discourse,” and “deficits” also occurred together at the network’s edges, which demonstrated that a wider variety of issues were investigated in CM studies. Despite these variations, it is safe to say that “metaphor,” “language,” and “comprehension” were essential and fundamental study issues, while “knowledge,” “mind,” and “discourse” had an impact on the development of CM and were also widely studied.

CM research trend

The bibliometric tool of Thematic Evolution and Thematic Trends is employed to predict the directions of potential future CM studies.

Thematic evolution

Examining Thematic Evolution and Trend Topics may reveal interesting research subjects and possible future orientations. Figure 13 demonstrates the dynamic nature of the metaphor study and the several research topics included. As time went by, “time,” “language,” and “mind” were maintained to be prominent academic areas. Metaphors may offer a practical and memorable method of structuring newly learned terminology. A lexical set is a concept that is well-known to most instructors. A linguistic set groups vocabulary according to a theme, such as “food” or “transportation.” By combining the words and sentences with a metaphorical meaning rather than a literal one, teachers may expand this concept to form “metaphorical sets.” Many scholars have shown their interest in this area. For example, Thomas and McRobbie (2001) emphasize how metaphors may help teachers and students establish a common language of learning. Andreou and Galantomos (2008) investigate the possibility of developing a conceptual curriculum for the instruction of metaphors and idioms in a foreign language setting.

Thematic evolution. Each hue represents a distinct subject, and the size of the rectangle means the depth of study.

However, throughout 2017–2022, “language” became the primary area of research interest, shifting attention away from “context” and other growing topics like “mind,” “comprehension,” “metaphor,” “discourse,” and “corpus” of CM. Some scholars conducted research into combining critical discourse analysis with self-constructed corpora of diverse genres to determine the underlying ideology metaphor ( Semino, 2008 ; Silvestre-López, 2020 ). Some academics concentrated on political speech and its associated discourse tactics. For instance, Pilyarchuk and Onysko (2018) found that Trump relied nearly entirely on conventional conceptual metaphors in his talks. Musolff (2011) studied the literary design of the dialog system in Shakespeare’s play and emphasized the general feature of metaphor’s dialogic role, which was further explored concerning the current use of body-based metaphor in political discourse. Koller (2004) explored metaphor and gender in electronic text corpora in the context of the commercial conversation. As for economic discourse, Chen (2018) utilized Wmatrix as a retrieval tool in conjunction with the “MIPVU” to identify and summarize the most prominent conceptual metaphors in economic speech and investigated the significance of metaphors.

Regarding discourse analysis, several efforts emphasized CM used in literary discourse ( Zhao and Zhou, 2019 ). Using the corpus tool Antconc3.2.4w, Zhao et al. (2020) conducted a study on Pearl S. Buck’s novel Dragon Seed and pointed out CMT and CBT were concerned with interpreting higher-order concepts such as meaning, language, sign, and representation and their interrelations. They complemented each other and contributed to discourse analysis. CM in literary works might be related to the writer’s cognitive and social contexts. Pearl Buck’s metaphorical thinking was closely related to her experiences in China. It may be extrapolated that these themes have a significant potential for CM research to continue to flourish.

Trend topics

Figure 14 indicates that, from 2002 to 2013, research subjects were relatively few, but their diversity increased after 2013. The wider the circle in the image, the greater the topic’s popularity among researchers was. Figure 14 shows that 2016 was a banner year for research on “metaphor,” “comprehension,” “discourse,” “mind,” and “metonymy,” as evidenced by the magnitude of the blue node. Between 2002 and 2022, “metaphor” was the most popular subject, appearing 143 times, followed by “language” (129), “comprehension” (61), “synthesis” (61), “metaphors” (43), “English” (38), “discovery” (33), “mind” (33), “meteorology” (28), “idioms” (26), and “space” (25). In 2016, research subjects were the most prevalent and featured the most often. They have been shown, once again, to be central and essential to CM research in recent years, and they may get even more emphasis in the years to come. It happened simultaneously as the Thematic Evolution, which ran from 2017 to 2022. In addition, critical new areas of study, including “metaphor,” “comprehension,” “discourse,” “corpus,” “brain,” “language,” and “mind,” maintained their popularity. The “corpus” of CM studies peaked in 2018 and predictions for its continual fruitfulness in the future were promising. Based on broad corpora, the first kind of investigation establishes the systematicity of conceptual metaphors or summarizes grammatical aspects that conventional metaphor studies cannot notice, compensating for CMT’s deficiencies ( Skorczynska and Deignan, 2006 ). Using CMT as an example, Charteris-Black (2004) proposed a novel research technique that integrated corpus linguistics, critical discourse analysis, and metaphor study to initiate a corpus-based metaphor study and develop new tools for identifying metaphors. Therefore, it is safe to say that “simile,” “adults,” “cancer,” “metaphors,” “words,” “brain,” “corpus,” and “perception” all have promising futures as research areas of CM.

Trend topics by keywords plus. The size of the nodes in Trend Topics represents the total number of publications for a particular topic and the peak popularity of that topic over time.

Conclusion and implication

This study employed a bibliometric technique to investigate 1,257 papers on CM research over the past two decades. The following are significant findings with productivity, content, and citation analysis. First, CM is a cognitive concept and has a widespread academic interest. “Metaphors,” “place,” “discourse,” and “corpus” were the central issues among the various study subjects. “Conceptual integration,” “comprehension,” “language,” and “mind” are also active and popular study topics in CM research. Second, in the past two decades, CM has been a research focus that has included many aspects, including authors, institutions, countries, and sources. Most of the cooperation survey was done with writers and institutions from many nations. The top five countries are Spain, the United States of America, China, Great Britain, and Russia. A rise in the number of academics studying CM suggests that CM research in cognitive linguistics applies to several facets of human cognition. Great Britain and China contributed the most to the growth of CM research, and substantial cooperation and networks were developed among them. These components of CM research are intertwined since the most cited individual contributes to establishing nations, institutions, and papers that significantly impact CM research.

Third, based on an examination of the Thematic Evolution and Trend Topic, we can infer the essential themes in CM research, such as “metaphors,” “discourse,” “space,” and “corpus,” may get greater attention in the future, which aligns with the Thematic Evolution between 2017 and 2022. In addition, “simile,” “adults,” “cancer,” “metaphors,” “words,” “brain,” “corpus,” “perception,” “conceptual integration,” “mind,” and “comprehension” will remain popular themes. The “interdisciplinarity” of CM demonstrates the effect of cognitive context, social context, and other cultural aspects on the framework of CM. The growing number of papers using bibliometric analysis across all disciplines suggests that it meets the desire of researchers who want proper research based on a wealth of literature.

This study will be helpful for beginners in the CM field, allowing them to classify information and find research results of CM quickly so that they may start their research projects. In addition, it may serve as a reference for seasoned researchers to comprehend the progress of CM research over the last two decades, find a suitable collaborator for their present research, and identify research gaps that they may block up in the future.

This study emphasizes the presentation of images and statistics because it is a quantitative study using a bibliometric tool based on data gathered from a database. However, it needs to go more in-depth to complete an evaluation of any specific theme of CM. We urge future research to broaden the study to use a range of more data gathering to examine concerns in CM to create a more thorough comprehension of CM.

Data availability statement

Author contributions.

XZ initiated the research idea, instructed YZ to analyze the data using bibliometric software, and co-wrote the article. Under the direction of XZ, YZ gathered and extracted the data and co-wrote the article’s analysis. XCZ contributed to the manuscript’s design, drafting the first part, introduction. All authors participated in revising and approving the version that was submitted.

Acknowledgments

We appreciate the reviewers for their insightful comments and ideas regarding the previous manuscript of this article. We also thank them for their excellent assistance in revising this article. Their insightful and thought-provoking remarks have at times, been difficult to respond to but have been vital to achieving the final form of this article.

1 https://www.bibliometrix.org/home

2 https://clarivate.com/webofsciencegroup/

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

• Abdul M. N., Syafiq, Ya Shak M., Mohamad F., Joharry S. A. (2022). Corpus-based studies of metaphor: An overview. Arab. World Engl. J. 13 512–528. 10.24093/awej/vol13no2.36 [ CrossRef ] [ Google Scholar ]
• Allahmoradi M. (2018). A review about the concept of metaphorical aspects of annex buildings. Cogent Eng. 5 :1501865. 10.1080/23311916.2018.1501865 [ CrossRef ] [ Google Scholar ]
• Allen D. (2006). Who’s in and who’s out? Language and the integration of new immigrant youth in Quebec. Int. J. Incl. Educ. 10 251–263. 10.1080/13603110500256103 [ CrossRef ] [ Google Scholar ]
• Alnajem M., Mostafa M. M., ElMelegy A. R. (2021). Mapping the first decade of circular economy research: A bibliometric network analysis. J. Ind. Prod. Eng. 38 29–50. 10.1080/21681015.2020.1838632 [ CrossRef ] [ Google Scholar ]
• Amirbagheri K., Núñez-Carballosa A., Guitart-Tarrés L., Merigó J. M. (2019). Research on green supply chain: A bibliometric analysis. Clean Technol. Environ. Policy 21 3–22. 10.1007/s10098-018-1624-1 [ CrossRef ] [ Google Scholar ]
• Andreou G., Galantomos L. (2008). Designing a conceptual syllabus for teaching metaphors and idioms in a foreign language context . Portalinguarum 9 , 69–78. [ Google Scholar ]
• Aria M., Cuccurullo C. (2017). Bibliometrix: An R-tool for comprehensive science mapping analysis . J. Inform . 11 , 959–975. 10.1016/j.joi.2017.08.007 [ CrossRef ] [ Google Scholar ]
• Bashir M. F., Ma B., Komal B., Bashir M. A. (2021). Analysis of environmental taxes publications: A bibliometric and systematic literature review. Environ. Sci. Pollut. Res. 28 20700–20716. 10.1007/s11356-020-12123-x [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bearman M., Mahoney P., Tai J., Castanelli D., Watling C. (2021). Invoking culture in medical education research: A critical review and metaphor analysis. Med. Educ. 55 903–911. 10.1111/medu.14464 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Birkle C., Pendlebury D. A., Schnell J., Adams J. (2020). Web of Science as a data source for research on scientific and scholarly activity. Quant. Sci. Stud. 1 363–376. 10.1162/qss_a_00018 [ CrossRef ] [ Google Scholar ]
• Bosman N., Taljard E. (2021). A cross-linguistic study of blood metaphors in Afrikaans and Northern Sotho. Lang. Matters 52 3–29. 10.1080/10228195.2020.1854331 [ CrossRef ] [ Google Scholar ]
• Bui T. D., Ali M. H., Tsai F. M., Iranmanesh M., Tseng M. L., Lim M. K. (2020). Challenges and trends in sustainable corporate finance: A bibliometric systematic review. J. Risk Financ. Manag. 13 :264. 10.3390/jrfm13110264 [ CrossRef ] [ Google Scholar ]
• Bundgaard P. F. (2019). The structure of our concepts: A critical assessment of conceptual metaphor theory as a theory of concepts. Cogn. Semiotics 12 1–11. 10.1515/cogsem-2019-2010 [ CrossRef ] [ Google Scholar ]
• Burgers C., Ahrens K. (2018). Change in metaphorical framing: Metaphors of trade in 225 years of state of the union addresses (1790–2014). Appl. Linguist. 41 260–279. 10.1093/applin/amy055 [ CrossRef ] [ Google Scholar ]
• Cameron L., Deignan A. (2006). The emergence of metaphor in discourse. Appl. Linguist. 27 671–690. 10.1093/applin/aml032 [ CrossRef ] [ Google Scholar ]
• Charteris-Black J. (2000). Metaphor and vocabulary teaching in ESP economics. English Specif. Purp. 19 149–165. 10.1016/S0889-4906(98)00025-8 [ CrossRef ] [ Google Scholar ]
• Charteris-Black J. (2004). Corpus approaches to critical metaphor analysis. Basingstoke: Palgrave Macmillan; 10.1057/9780230000612 [ CrossRef ] [ Google Scholar ]
• Chen L. (2018). A study on the metaphor of english economic discourse based on corpus. J. Plaster Inst. Foreign Chin. 41 34–42. [ Google Scholar ]
• Chiu W. T., Ho Y. S. (2007). Bibliometric analysis of Tsunami research. Scientometrics 73 3–17. 10.1007/s11192-005-1523-1 [ CrossRef ] [ Google Scholar ]
• Cuccio V., Steen G. (2019). Deliberate metaphors and embodied simulation , in Current approaches to metaphor analysis in discourse , ed Ignasi N. F. (Berlin: De Gruyter Mouton; ), 185–204. 10.1515/9783110629460-009 [ CrossRef ] [ Google Scholar ]
• Cuccio V., Perconti P., Steen G., Shtyrov Y., Huang Y. (2022). Editorial: Experimental approaches to pragmatics. Front. Psychol. 13 :865737. 10.3389/fpsyg.2022.865737 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Deignan A., Potter L. (2004). A corpus study of metaphors and metonyms in English and Italian. J. Pragmat. 36 1231–1252. 10.1016/j.pragma.2003.10.010 [ CrossRef ] [ Google Scholar ]
• Donthu N., Kumar S., Mukherjee D., Pandey N., Lim W. M. (2021). How to conduct a bibliometric analysis: An overview and guidelines. J. Bus. Res. 133 285–296. 10.1016/j.jbusres.2021.04.070 [ CrossRef ] [ Google Scholar ]
• Efron N., Morgan P. B., Jones L. W., Nichols J. J. (2021). Bibliometric analysis of the refractive error field. Clin. Exp. Optom. 104 641–643. 10.1080/08164622.2021.1880868 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Ellegaard O., Wallin J. A. (2015). The bibliometric analysis of scholarly production: How great is the impact? Scientometrics 105 1809–1831. 10.1007/s11192-015-1645-z [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fabiszak M. (2007). A conceptual metaphor approach to war discourse and its implications. Poznan: Adam Mickiewicz University Press. [ Google Scholar ]
• Fahimnia B., Sarkis J., Davarzani H. (2015). Green supply chain management: A review and bibliometric analysis. Int. J. Prod. Econ. 162 101–114. [ Google Scholar ]
• Fauconnier G., Turner M. (1998). Conceptual integration networks. Cogn. Sci. 22 133–187. 10.1207/s15516709cog2202_1 [ CrossRef ] [ Google Scholar ]
• Faust O., Hagiwara Y., Hong T. J., Lih O. S., Acharya U. R. (2018). Deep learning for healthcare applications based on physiological signals: A review. Comput. Methods Programs Biomed. 161 1–13. 10.1016/j.cmpb.2018.04.005 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Ferrari F. (2007). Metaphor at work in the analysis of political discourse: Investigating a “preventive war” persuasion strategy. Discourse Soc. 18 603–625. 10.1177/0957926507079737 [ CrossRef ] [ Google Scholar ]
• Gandolfo S. (2019). Metaphors of metaphors: Reflections on the use of conceptual metaphor theory in premodern Chinese texts. Dao J. Comp. Philos. 18 323–345. 10.1007/s11712-019-09669-0 [ CrossRef ] [ Google Scholar ]
• Geissdoerfer M., Savaget P., Bocken N. M. P., Hultink E. J. (2017). The circular economy–A new sustainability paradigm? J. Clean. Prod. 143 757–768. 10.1016/j.jclepro.2016.12.048 [ CrossRef ] [ Google Scholar ]
• Gentner D., Imai M., Boroditsky L. (2002). As time goes by: Evidence for two systems in processing space → time metaphors. Lang. Cogn. Process. 17 537–565. 10.1080/01690960143000317 [ CrossRef ] [ Google Scholar ]
• Georgescu S. (2020). Conceptual metaphors leading to some names of anger in the Indo-European languages (with focus on the romance languages). Revue Roumaine Delinguistique-Romanian Rev. Linguist. 65 189–197. [ Google Scholar ]
• Gibbs R. W., Jr. (2006). Metaphor interpretation as embodied simulation . Mind Lang . 21 , 434–458. [ Google Scholar ]
• Gibbs R. W. (2013). The real complexities of psycholinguistic research on metaphor. Lang. Sci. 40 45–52. 10.1016/j.langsci.2013.03.001 [ CrossRef ] [ Google Scholar ]
• Gibbs R. W., Costa Lima P. L., Francozo E. (2004). Metaphor is grounded in embodied experience. J. Pragmat. 36 1189–1210. 10.1016/j.pragma.2003.10.009 [ CrossRef ] [ Google Scholar ]
• Giora R. (2002). Literal vs. Figurative language: Different or equal? J. Pragmat. 34 487–506. 10.1016/S0378-2166(01)00045-5 [ CrossRef ] [ Google Scholar ]
• Greenacre M., Blasius J. (2006). Multiple correspondence analysis and related methods. New York, NY: Chapman and Hall/CRC. 10.1201/9781420011319 [ CrossRef ] [ Google Scholar ]
• Hashemi H., Rajabi R., Brashear-Alejandro T. G. (2022). COVID-19 research in management: An updated bibliometric analysis. J. Bus. Res. 149 795–810. 10.1016/j.jbusres.2022.05.082 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hasson U., Glucksberg S. (2006). Does understanding negation entail affirmation? An examination of negated metaphors. J. Pragmat. 38 1015–1032. 10.1016/j.pragma.2005.12.005 [ CrossRef ] [ Google Scholar ]
• Herold E., Pöckelmann M., Berg C., Ritter J., Hall M. M. (2019). “ Stable word-clouds for visualising text-changes over time ,” in International conference on theory and practice of digital libraries , eds Antoine D., Antoine I., Golub G., Trond A., Adam J. (Cham: Springer; ), 224–237. 10.1007/978-3-030-30760-8_20 [ CrossRef ] [ Google Scholar ]
• Holyoak K. J., Stamenkovic D. (2018). Metaphor comprehension: A critical review of theories and evidence. Psychol. Bull. 144 :641. 10.1037/bul0000145 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jain N., Virmani D., Abraham A. (2021). Tsunami in the last 15 years: A bibliometric analysis with a detailed overview and future directions. Nat. Hazards 106 139–172. 10.1007/s11069-020-04454-2 [ CrossRef ] [ Google Scholar ]
• Jensen L. X., Bearman M., Boud D. (2021). Understanding feedback in online learning–a critical review and metaphor analysis. Comput. Educ. 173 :104271. 10.1016/j.compedu.2021.104271 [ CrossRef ] [ Google Scholar ]
• Kazemian R., Hatamzadeh S. (2022). COVID-19 in english and persian: A cognitive linguistic study of illness metaphors across languages. Metaphor. Symb. 37 152–170. 10.1080/10926488.2021.1994839 [ CrossRef ] [ Google Scholar ]
• Koller V. (2004). Metaphor and gender in business media discourse a critical cognitive study. New York, NY: Palgrave Macmillan. 10.1057/9780230511286 [ CrossRef ] [ Google Scholar ]
• Kövecses Z. (2020). An extended view of conceptual metaphor theory . Rev. Cogn. Ling . 18 , 112–130. [ Google Scholar ]
• Lakoff G. (1987). Women, fire and dangerous things: What categories reveal about the mind. Chicago, IL: The University of Chicago Press. 10.7208/chicago/9780226471013.001.0001 [ CrossRef ] [ Google Scholar ]
• Lakoff G., Johnson M. (1980). Metaphors we live by. Chicago, IL: The University of Chicago Press. [ Google Scholar ]
• Lakoff G., Johnson M. (2003). Metaphors we live by . Chicago, IL: The University of Chicago Press. [ Google Scholar ]
• Lakoff G., Johnson M. (2008). Metaphors we live by. Chicago, IL: University of Chicago press. [ Google Scholar ]
• Lakoff G., Turner M. (1989). More than cool reason: A field guide to poetic metaphor. Chicago, IL: University of Chicago Press. [ Google Scholar ]
• Littlemore J., Low G. (2006). Metaphoric competence, second language learning, and communicative language ability. Appl. Linguist. 27 268–294. [ Google Scholar ]
• Luis E. C., Celma D. (2020). Circular economy. A review and bibliometric analysis. Sustainability 12 :6381. [ Google Scholar ]
• Maestre M. D. L. (2020). Gender, ideology and conceptual metaphors: Women and the source domain of the hunt. J. Engl. Stud. 28 203–218. [ Google Scholar ]
• Murphy G. L. (1996). On metaphoric representation. Urbana, IL: University of Illinois. [ PubMed ] [ Google Scholar ]
• Musolff A. (2011). Metaphor in political dialogue. Lang. Dialogue 1 191–206. [ Google Scholar ]
• Musolff A. (2012). The study of metaphor as part of critical discourse analysis. Crit. Discourse Stud. 9 301–310. [ Google Scholar ]
• Pilyarchuk K., Onysko A. (2018). Conceptual metaphors in Donald Trump’s political speeches: Framing his topics and (self-) constructing his persona. Colloquium: New Philologies 3 , 98–156. [ Google Scholar ]
• Prandi M. (2017). Conceptual conflicts in metaphors and figurative language. New York, NY: Routledge. [ Google Scholar ]
• Qiu H., Watson B., Tay D. (2022). Metaphors and trauma: An image schematic analysis of symptom-specific metaphors. Lingua 271 :103244. [ Google Scholar ]
• Rizzato I. (2021). Conceptual conflicts in metaphors and pragmatic strategies for their translation. Front. Psychol. 12 :662276. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rizzato I. (2022). “ Shakespeare’s metaphorical swarms in the two gentlemen of verona: Textual and pragmatic features and their impact on the Italian translation ,” in Researching metaphors , eds Michele P., Micaela R. (Abingdon: Routledge; ), 245–259. [ Google Scholar ]
• Saheb T., Saheb T., Carpenter D. O. (2021). Mapping research strands of ethics of artificial intelligence in healthcare: A bibliometric and content analysis. Comput. Biol. Med. 135 :104660. [ PubMed ] [ Google Scholar ]
• Sarkodie S. A., Strezov V. (2019). A review on Environmental Kuznets Curve hypothesis using bibliometric and meta-analysis. Sci. Total Environ. 649 128–145. [ PubMed ] [ Google Scholar ]
• Schaab G., Adams S., Coetzee S. (2022). Conveying map finesse: Thematic map making essentials for today’s university students. J. Geogr. High. Educ. 46 101–127. [ Google Scholar ]
• Semino E. (2002). A study baby or a derailing train? Metaphorical representations of the euro in British and Italian newspapers. Text & Talk 22 , 107–139. [ Google Scholar ]
• Semino E. (2008). Metaphor in discourse. Cambridge, MA: Cambridge University Press. [ Google Scholar ]
• Shutova E., Devereux B. J., Korhonen A. (2013). Conceptual metaphor theory meets the data: A corpus-based human annotation study. Lang. Resour. Eval. 47 1261–1284. [ Google Scholar ]
• Silvestre-López A. J. (2020). Conceptual metaphor in meditation discourse: An analysis of the spiritual perspective. GEMA Online J. Lang. Stud. 20 35–53. [ Google Scholar ]
• Sinclair J. (1991). Corpus, concordance, collocation. Oxford: Oxford University Press. [ Google Scholar ]
• Skorczynska H., Deignan A. (2006). Readership and purpose in the choice of economics metaphors. Metaphor Symb. 21 87–104. [ Google Scholar ]
• Stamenković D., Ichien N., Holyoak K. J. (2019). Metaphor comprehension: An individual-differences approach. J. Mem. Lang. 105 108–118. [ Google Scholar ]
• Thomas G. P., McRobbie C. J. (2001). Using a metaphor for learning to improve students’ metacognition in the chemistry classroom. J. Res. Sci. Teach. 38 222–259. [ Google Scholar ]
• Tissari H. (2010). “ Love, metaphor, and responsibility: Some examples from early modern and present-day english corpora ,” in Researching and applying metaphor in the real world , eds Low G., Todd Z., Deignan A., Cameron L. (Amsterdam: John Benjamins Publishing Company; ). [ Google Scholar ]
• Tohidian I., Rahimian H. (2019). Bringing Morgan’s metaphors in organization contexts: An essay review. Cogent Bus. Manag. 6 1–8. [ Google Scholar ]
• Verma S., Gustafsson A. (2020). Investigating the emerging COVID-19 research trends in the field of business and management: A bibliometric analysis approach. J. Bus. Res. 118 253–261. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Waltman L., van Eck N. J., Noyons E. C. M. (2010). A unified approach to mapping and clustering of bibliometric networks. J. Informetr. 4 629–635. [ Google Scholar ]
• Yan P., Li M., Li J., Lu Z., Hui X., Bai Y., et al. (2021). Bibliometric analysis and systematic review of global coronavirus research trends before COVID-19: Prospects and implications for COVID-19 research. Front. Med. 8 :729138. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Zhao X. (2008). Restriction of cognitive context in the construal of metaphor. Foreign Lang. Teach. 9 , 22–24+33. [ Google Scholar ]
• Zhao X., Zhou W. (2019). A corpus-based explanation of metaphors in A House Divided. Lang. Semiotic Stud. 5 108–127. [ Google Scholar ]
• Zhao X., Han Y. Y., Zhao X. C. (2019). A corpus-based study of metaphor in pavilion of women. Chin. Semiotic Stud. 15 , 95–117. 10.1515/css-2019-0006 [ CrossRef ] [ Google Scholar ]
• Zhao X., Shen R., Zhao X. C. (2020). A cognitive-semiotic construal of metaphor in discourse: A corpus-based approach. Chin. Semiotic Stud. 16 , 119–143. 10.1515/css-2020-0006 [ CrossRef ] [ Google Scholar ]
• Zhu S., Liu Y., Gu Z., Zhao Y. (2021). A bibliometric analysis of advanced healthcare materials: Research trends of biomaterials in healthcare application. Adv. Healthc. Mater. 10 :2002222. [ PubMed ] [ Google Scholar ]
• Zupic I., Čater T. (2015). Bibliometric methods in management and organization. Organ. Res. Methods 18 429–472. [ Google Scholar ]

Help | Advanced Search

Computer Science > Computation and Language

Title: leave no context behind: efficient infinite context transformers with infini-attention.

Abstract: This work introduces an efficient method to scale Transformer-based Large Language Models (LLMs) to infinitely long inputs with bounded memory and computation. A key component in our proposed approach is a new attention technique dubbed Infini-attention. The Infini-attention incorporates a compressive memory into the vanilla attention mechanism and builds in both masked local attention and long-term linear attention mechanisms in a single Transformer block. We demonstrate the effectiveness of our approach on long-context language modeling benchmarks, 1M sequence length passkey context block retrieval and 500K length book summarization tasks with 1B and 8B LLMs. Our approach introduces minimal bounded memory parameters and enables fast streaming inference for LLMs.

Submission history

Access paper:.

• HTML (experimental)
• Other Formats

References & Citations

• Google Scholar
• Semantic Scholar

BibTeX formatted citation

Bibliographic and Citation Tools

Code, data and media associated with this article, recommenders and search tools.

• Institution

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .

Francis Collins: Why I’m going public with my prostate cancer diagnosis

I served medical research. now it’s serving me. and i don’t want to waste time..

Over my 40 years as a physician-scientist, I’ve had the privilege of advising many patients facing serious medical diagnoses. I’ve seen them go through the excruciating experience of waiting for the results of a critical blood test, biopsy or scan that could dramatically affect their future hopes and dreams.

But this time, I was the one lying in the PET scanner as it searched for possible evidence of spread of my aggressive prostate cancer . I spent those 30 minutes in quiet prayer. If that cancer had already spread to my lymph nodes, bones, lungs or brain, it could still be treated — but it would no longer be curable.

Why am I going public about this cancer that many men are uncomfortable talking about? Because I want to lift the veil and share lifesaving information, and I want all men to benefit from the medical research to which I’ve devoted my career and that is now guiding my care.

Five years before that fateful PET scan, my doctor had noted a slow rise in my PSA, the blood test for prostate-specific antigen. To contribute to knowledge and receive expert care, I enrolled in a clinical trial at the National Institutes of Health, the agency I led from 2009 through late 2021.

At first, there wasn’t much to worry about — targeted biopsies identified a slow-growing grade of prostate cancer that doesn’t require treatment and can be tracked via regular checkups, referred to as “active surveillance.” This initial diagnosis was not particularly surprising. Prostate cancer is the most commonly diagnosed cancer in men in the United States, and about 40 percent of men over age 65 — I’m 73 — have low-grade prostate cancer . Many of them never know it, and very few of them develop advanced disease.

Why am I going public about this cancer that many men are uncomfortable talking about? Because I want to lift the veil and share lifesaving information.

But in my case, things took a turn about a month ago when my PSA rose sharply to 22 — normal at my age is less than 5. An MRI scan showed that the tumor had significantly enlarged and might have even breached the capsule that surrounds the prostate, posing a significant risk that the cancer cells might have spread to other parts of the body.

New biopsies taken from the mass showed transformation into a much more aggressive cancer. When I heard the diagnosis was now a 9 on a cancer-grading scale that goes only to 10, I knew that everything had changed.

Thus, that PET scan, which was ordered to determine if the cancer had spread beyond the prostate, carried high significance. Would a cure still be possible, or would it be time to get my affairs in order? A few hours later, when my doctors showed me the scan results, I felt a rush of profound relief and gratitude. There was no detectable evidence of cancer outside of the primary tumor.

Later this month, I will undergo a radical prostatectomy — a procedure that will remove my entire prostate gland. This will be part of the same NIH research protocol — I want as much information as possible to be learned from my case, to help others in the future.

While there are no guarantees, my doctors believe I have a high likelihood of being cured by the surgery.

My situation is far better than my father’s when he was diagnosed with prostate cancer four decades ago. He was about the same age that I am now, but it wasn’t possible back then to assess how advanced the cancer might be. He was treated with a hormonal therapy that might not have been necessary and had a significant negative impact on his quality of life.

Because of research supported by NIH, along with highly effective collaborations with the private sector, prostate cancer can now be treated with individualized precision and improved outcomes.

As in my case, high-resolution MRI scans can now be used to delineate the precise location of a tumor. When combined with real-time ultrasound, this allows pinpoint targeting of the prostate biopsies. My surgeon will be assisted by a sophisticated robot named for Leonardo da Vinci that employs a less invasive surgical approach than previous techniques, requiring just a few small incisions.

Advances in clinical treatments have been informed by large-scale, rigorously designed trials that have assessed the risks and benefits and were possible because of the willingness of cancer patients to enroll in such trials.

I feel compelled to tell this story openly. I hope it helps someone. I don’t want to waste time.

If my cancer recurs, the DNA analysis that has been carried out on my tumor will guide the precise choice of therapies. As a researcher who had the privilege of leading the Human Genome Project , it is truly gratifying to see how these advances in genomics have transformed the diagnosis and treatment of cancer.

I want all men to have the same opportunity that I did. Prostate cancer is still the No. 2 cancer killer among men. I want the goals of the Cancer Moonshot to be met — to end cancer as we know it. Early detection really matters, and when combined with active surveillance can identify the risky cancers like mine, and leave the rest alone. The five-year relative survival rate for prostate cancer is 97 percent, according to the American Cancer Society , but it’s only 34 percent if the cancer has spread to distant areas of the body.

But lack of information and confusion about the best approach to prostate cancer screening have impeded progress. Currently, the U.S. Preventive Services Task Force recommends that all men age 55 to 69 discuss PSA screening with their primary-care physician, but it recommends against starting PSA screening after age 70.

Other groups, like the American Urological Association , suggest that screening should start earlier, especially for men with a family history — like me — and for African American men, who have a higher risk of prostate cancer. But these recommendations are not consistently being followed.

Our health-care system is afflicted with health inequities. For example, the image-guided biopsies are not available everywhere and to everyone. Finally, many men are fearful of the surgical approach to prostate cancer because of the risk of incontinence and impotence, but advances in surgical techniques have made those outcomes considerably less troublesome than in the past. Similarly, the alternative therapeutic approaches of radiation and hormonal therapy have seen significant advances.

A little over a year ago, while I was praying for a dying friend, I had the experience of receiving a clear and unmistakable message. This has almost never happened to me. It was just this: “Don’t waste your time, you may not have much left.” Gulp.

Having now received a diagnosis of aggressive prostate cancer and feeling grateful for all the ways I have benefited from research advances, I feel compelled to tell this story openly. I hope it helps someone. I don’t want to waste time.

Francis S. Collins served as director of the National Institutes of Health from 2009 to 2021 and as director of the National Human Genome Research Institute at NIH from 1993 to 2008. He is a physician-geneticist and leads a White House initiative to eliminate hepatitis C in the United States, while also continuing to pursue his research interests as a distinguished NIH investigator.

An earlier version of this article said prostate cancer is the No. 2 killer of men. It is the No. 2 cause of cancer death among men. The article has been updated.

• How to revive hair that thins, grays or gets out of control as you age April 15, 2024 How to revive hair that thins, grays or gets out of control as you age April 15, 2024
• Anxious about your first mammogram? Here’s what I learned. April 13, 2024 Anxious about your first mammogram? Here’s what I learned. April 13, 2024
• Medical Mysteries: Years of hives and fevers traced to a startling cause April 13, 2024 Medical Mysteries: Years of hives and fevers traced to a startling cause April 13, 2024