“For crying out loud, don't call me a warrior”: Standpoints of resistance against violence metaphors for cancer

Expressions of resistance to violence metaphors for cancer provide important insights as to how these metaphors are perceived and which specific aspects of their use are deemed inapt or inappropriate by actual language users. The first objective of this paper is to examine different types of resistance standpoints that are expressed in critical responses to violence metaphors for cancer as well as the different metaphor dimensions these standpoints are focused on. Secondly, the paper aims to demonstrate the value of a close analysis of standpoints of resistance to metaphor from a combined argumentation theoretical and metaphor theoretical perspective. It will be shown how a combined argumentation theoretical and metaphor theoretical framework constitutes a useful and structured method for understanding the nature of resistance standpoints and the specific dimension(s) of metaphor they address. In this paper the combined framework will be put into practice in six case studies, each illustrating different manners in which language users express their disapproval of a particular use of violence metaphors for cancer. © 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1. Resistance to violence metaphors for cancer The use of violencemetaphors for cancer has been a subject of discussion for over three decades (Fleischman, 2008). These decades have seen a rise in academic studies on the (potential) shortcomings of these metaphors, which has resulted in important findings on their different functions and effects (e.g. Semino et al., 2015; Hendricks et al., 2018). One important aspect that has failed to receive attention in these many studies on violence metaphors for cancer, however, is the fact that language users can be found to express conflicting points of view about the aptness or appropriateness of these metaphors: language users’ resistance to violence metaphors for cancer has prompted many a discussion in newspapers, magazines, internet forums and blogs (see for instance Granger, 2014). Public discussions on violence metaphors for cancer provide a wealth of data about the negative points of view language users hold towards these metaphors; a close analysis of such discussions provides valuable insights about which aspects of these metaphors are deemed inapt or inappropriate, and why. More particularly, examining standpoints of resistance that are taken in public discussions on the use of violence metaphors for cancer can help us pinpoint what language users find problematic about (a) (given) metaphor(s) and which conclusion they draw about the metaphor’(s)(’) use. Examples of such standpoints of resistance include expressions such as “People should stop encouraging patients to ‘fight bravely’”, “Violence metaphors put the burden of healing on cancer patients” and “I dislike seeing myself as a ‘cancer warrior’”, which differ from one another in a number of ways. Wackers), H.J.Plug@uva.nl (H.J. Plug), G.J.Steen@uva.nl (G.J. Steen). r B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4. D.Y.M. Wackers, H.J. Plug and G.J. Steen Journal of Pragmatics 174 (2021) 68e77 This study aims to provide a better understanding of different resistance standpoints in terms of the nature of the proposition they express and the dimension of metaphor use that forms the target of resistance. The study also aims to demonstrate how a combination of theoretical analytical tools from pragma-dialectical argumentation theory (Van Eemeren et al., 2008) and Deliberate Metaphor Theory (DMT) (Steen, 2015) constitutes a useful and structured method for a close analysis of resistance standpoints that takes into account characteristics of the resistance standpoint as well as properties of the contestedmetaphors. The combined argumentation andmetaphor theoretical approach taken in this studywill be argued to render a more encompassing and more detailed account of differences of opinion on violence metaphors for cancer compared to an exclusively argumentation theoretical or exclusivelymetaphor theoretical approach: An analysis of resistance to metaphor from a metaphor theoretical perspective alone fails to capture the different positions language users hold regarding a contested metaphor; an analysis of resistance to metaphor that is solely based on argumentation theoretical insights lacks the instruments necessary to pinpoint the dimensions of metaphor expressions of resistance are focused at.1 We will discuss the theoretical framework underpinning our analysis of metaphor (DMT) and the approach upon which our argumentative analysis of resistance standpoints is based (Pragma-dialectics) in more detail in Section 2 of this paper. In Section 3 wewill put both theories into practice and discuss six examples of resistance standpoints that differ in terms of the type of proposition they express and the dimension of metaphor they address. More specifically, through these case studies wewill demonstrate how resistance standpoints may address a) the role of metaphor in thinking about the metaphor's target domain, b) the role of metaphor in talking about the metaphor's target domain, and/or c) the role of metaphor asmetaphor in communication whereby distinct attention is paid to the source domain of violence as a separate domain of reference. These three propositional aspects correspond to the three dimensions of metaphor that are distinguished within DMT, i.e., the conceptual, linguistic, and communicative dimension of metaphor use. Furthermore, in line with pragma-dialectical theory, in this paper three types of resistance standpoints will be distinguished: evaluative, descriptive, and prescriptive standpoints. The pragma-dialectical distinction between these types will help to understand not only how violence metaphors for cancer are negatively perceived in public discourse, but also the different conclusions at which language users can arrive regarding the contested metaphor use. 2. Theoretical framework In principle, ‘resistance to metaphor’ may refer to any form of opposition to the use of metaphor in a given context. In the current studywe are specifically interested in argumentative resistance, or resistance that is supported byarguments.We seek to understand the various negative positions language users take inpublic discussions on the use of violencemetaphors for cancer. That is to say, refusing to pick up on ametaphor used by someone else could be considered an example of resistance tometaphor too; this and other forms of criticism that are not expressed verbally by means of (counter)argumentation will be left out of consideration here. We will give a more detailed explanation of what we mean by ‘standpoints of resistance to violence metaphors for cancer’ below,wherewewill discuss the theoretical foundations of DMT (Subsection 2.1) andPragma-dialectics (2.2). 2.1. Deliberate Metaphor Theory Metaphor can be defined as ameans to talk and think about one thing in terms of another due to some perceived similarity between these two things (Semino, 2008, p. 1). This definition reflects the cognitive-linguistic distinction between metaphor in language (‘talk’) and thought (‘think’). In the past decades, much cognitive-linguistic research on metaphor has been conductedwithin the framework of Conceptual Metaphor Theory (CMT) (Lakoff and Johnson,1980). One of CMT's main tenets is that metaphorical expressions in language are surface phenomena that derive from conceptual structures in thought.When we talk about cancer as war, for instance, this is due to our understanding of (overcoming) cancer as sharing similarities with (waging) war.2 Even though the two concepts belong to different semantic domains, our mind perceives the structural relationships they share. These relationships are also referred to as the ‘cross-domain mappings’ brought about by metaphor. The cognitive-linguistic approach has had a profound impact on metaphor theory and research. Recent years, however, have seen an increase in more rhetorically-oriented approaches to metaphor taking into account another aspect of metaphor (i.a. Cameron, 2003; Charteris-Black and Musolff, 2003; Steen, 2015; for an extensive overview see Reijnierse, 2017, pp. 4e7). These approaches consider metaphor not only as a linguistic phenomenon that can be connected to conceptual structures in thought, but also as a communicative means employed in interaction between language users. DMT (Steen, 2015) recognizes the communicative function of metaphor as a separate, third dimension of metaphor e i.e., in addition to the dimensions of metaphor in thought and metaphor in language. The theory argues that every metaphor can be characterised along each of the three dimensions. In thought, metaphors are based on an understanding of one conceptual 1 For recent research that combines insights from DMT with Pragma-dialectics in the analysis of metaphor use in argumentation, see Van Poppel (2020). 2 For our definition of violence metaphors for cancer we make use of Demmen et al.’s characterisation of violence metaphors as: “any metaphorical expressions or similes whose literal meanings suggest scenarios in which, prototypically, a human agent intentionally causes physical harm to another human, with or without weapons. Less prototypical scenarios involve non-human agents, the threat or consequences of violence, or non-physical harm.” (2015, pp. 211e212).


Resistance to violence metaphors for cancer
The use of violence metaphors for cancer has been a subject of discussion for over three decades (Fleischman, 2008). These decades have seen a rise in academic studies on the (potential) shortcomings of these metaphors, which has resulted in important findings on their different functions and effects (e.g. Semino et al., 2015;Hendricks et al., 2018). One important aspect that has failed to receive attention in these many studies on violence metaphors for cancer, however, is the fact that language users can be found to express conflicting points of view about the aptness or appropriateness of these metaphors: language users' resistance to violence metaphors for cancer has prompted many a discussion in newspapers, magazines, internet forums and blogs (see for instance Granger, 2014).
Public discussions on violence metaphors for cancer provide a wealth of data about the negative points of view language users hold towards these metaphors; a close analysis of such discussions provides valuable insights about which aspects of these metaphors are deemed inapt or inappropriate, and why. More particularly, examining standpoints of resistance that are taken in public discussions on the use of violence metaphors for cancer can help us pinpoint what language users find problematic about (a) (given) metaphor(s) and which conclusion they draw about the metaphor'(s)(') use. Examples of such standpoints of resistance include expressions such as "People should stop encouraging patients to 'fight bravely'", "Violence metaphors put the burden of healing on cancer patients" and "I dislike seeing myself as a 'cancer warrior'", which differ from one another in a number of ways.
This study aims to provide a better understanding of different resistance standpoints in terms of the nature of the proposition they express and the dimension of metaphor use that forms the target of resistance. The study also aims to demonstrate how a combination of theoretical analytical tools from pragma-dialectical argumentation theory (Van Eemeren et al., 2008) and Deliberate Metaphor Theory (DMT) (Steen, 2015) constitutes a useful and structured method for a close analysis of resistance standpoints that takes into account characteristics of the resistance standpoint as well as properties of the contested metaphors. The combined argumentation and metaphor theoretical approach taken in this study will be argued to render a more encompassing and more detailed account of differences of opinion on violence metaphors for cancer compared to an exclusively argumentation theoretical or exclusively metaphor theoretical approach: An analysis of resistance to metaphor from a metaphor theoretical perspective alone fails to capture the different positions language users hold regarding a contested metaphor; an analysis of resistance to metaphor that is solely based on argumentation theoretical insights lacks the instruments necessary to pinpoint the dimensions of metaphor expressions of resistance are focused at. 1 We will discuss the theoretical framework underpinning our analysis of metaphor (DMT) and the approach upon which our argumentative analysis of resistance standpoints is based (Pragma-dialectics) in more detail in Section 2 of this paper. In Section 3 we will put both theories into practice and discuss six examples of resistance standpoints that differ in terms of the type of proposition they express and the dimension of metaphor they address. More specifically, through these case studies we will demonstrate how resistance standpoints may address a) the role of metaphor in thinking about the metaphor's target domain, b) the role of metaphor in talking about the metaphor's target domain, and/or c) the role of metaphor as metaphor in communication whereby distinct attention is paid to the source domain of violence as a separate domain of reference. These three propositional aspects correspond to the three dimensions of metaphor that are distinguished within DMT, i.e., the conceptual, linguistic, and communicative dimension of metaphor use. Furthermore, in line with pragma-dialectical theory, in this paper three types of resistance standpoints will be distinguished: evaluative, descriptive, and prescriptive standpoints. The pragma-dialectical distinction between these types will help to understand not only how violence metaphors for cancer are negatively perceived in public discourse, but also the different conclusions at which language users can arrive regarding the contested metaphor use.

Theoretical framework
In principle, 'resistance to metaphor' may refer to any form of opposition to the use of metaphor in a given context. In the current study we are specifically interested in argumentative resistance, or resistance that is supported by arguments. We seek to understand the various negative positions language users take in public discussions on the use of violence metaphors for cancer. That is to say, refusing to pick up on a metaphor used by someone else could be considered an example of resistance to metaphor too; this and other forms of criticism that are not expressed verbally by means of (counter)argumentation will be left out of consideration here. We will give a more detailed explanation of what we mean by 'standpoints of resistance to violence metaphors for cancer' below, where we will discuss the theoretical foundations of DMT (Subsection 2.1) and Pragma-dialectics (2.2).

Deliberate Metaphor Theory
Metaphor can be defined as a means to talk and think about one thing in terms of another due to some perceived similarity between these two things (Semino, 2008, p. 1). This definition reflects the cognitive-linguistic distinction between metaphor in language ('talk') and thought ('think'). In the past decades, much cognitive-linguistic research on metaphor has been conducted within the framework of Conceptual Metaphor Theory (CMT) (Lakoff and Johnson, 1980). One of CMT's main tenets is that metaphorical expressions in language are surface phenomena that derive from conceptual structures in thought. When we talk about cancer as war, for instance, this is due to our understanding of (overcoming) cancer as sharing similarities with (waging) war. 2 Even though the two concepts belong to different semantic domains, our mind perceives the structural relationships they share. These relationships are also referred to as the 'cross-domain mappings' brought about by metaphor.
The cognitive-linguistic approach has had a profound impact on metaphor theory and research. Recent years, however, have seen an increase in more rhetorically-oriented approaches to metaphor taking into account another aspect of metaphor (i.a. Cameron, 2003;Charteris-Black and Musolff, 2003;Steen, 2015; for an extensive overview see Reijnierse, 2017, pp. 4e7). These approaches consider metaphor not only as a linguistic phenomenon that can be connected to conceptual structures in thought, but also as a communicative means employed in interaction between language users.
DMT (Steen, 2015) recognizes the communicative function of metaphor as a separate, third dimension of metaphor e i.e., in addition to the dimensions of metaphor in thought and metaphor in language. The theory argues that every metaphor can be characterised along each of the three dimensions. In thought, metaphors are based on an understanding of one conceptual domain in terms of another, e.g., CANCER (the metaphor's target domain) and VIOLENCE (the metaphor's source domain). In language use, metaphors take form of particular metaphorical expressions such as 'she fought a battle against cancer' (an indirect metaphor) or 'cancer is like war' (a simile). In communication between language users, metaphors do or do not "serve as metaphors" between sender and addressee (Steen, 2017, p.1). The metaphorical use of the preposition 'up' in "speak up", for instance, is ingrained in our language and may often go unnoticed as metaphor. The use of a novel nominal metaphor in "cancer researchers need a 'battlespace vision'", on the other hand, deviates from the typical way in which people talk and may potentially influence language users' understanding of the topic of communication in relation to the metaphor's source domain. If in a particular situation of use this is indeed the case and language users, in production or reception, pay distinct attention to the metaphor's source domain as a separate domain of reference, the metaphor would concern a deliberate metaphor .
Whether a given metaphor actually makes an addressee move away their attention from the metaphor's target domain to its source domain is something that can only be established by psycholinguistic and psychological research. DMT takes a strictly semiotic approach to the identification and analysis of metaphor in language use based on theories of text comprehension in discourse psychology (e.g., Van Dijk and Kintsch, 1983;McNamara and Magliano, 2009). 3 More specifically, DMT describes the meaning of metaphorical utterances in a structuralefunctional way and takes as a starting point the systematic and reliable identification of potentially deliberate metaphor in language use. According to the theory, a metaphor counts as potentially deliberate "when the source domain of the metaphor is part of the referential meaning of the utterance in which it is used" (Reijnierse et al., 2017, p. 134). 4 Most analyses that have been conducted within the framework of DMT have been concerned with metaphor use in text and transcripts of talk. In this paper we will apply the framework of DMT in a novel way. We will use DMT's differentiation between metaphor as a matter of thought, language, and communication as a starting point in our analysis of resistance standpoints. More specifically, we will argue that the three dimensions of metaphor cannot only be used to characterise metaphor use, but that each of these dimensions can potentially also be identified in critical responses to metaphor use e i.e., that in some expressions of resistance to violence metaphors for cancer particular dimensions of metaphor use can be found to play a more prominent role. More prominent, as each metaphor in language use has a linguistic, conceptual, as well as a communicative dimension, one of which can potentially be resisted more explicitly compared to the other dimensions. Resistance standpoints may, for instance, explicitly oppose the use of one or more source domain term(s) in conversations about a given target domain, i.e., focusing on the dimension of metaphor in language. But they can also, and secondly, specifically resist the use of (a) metaphor in thinking about a target domain, the dimension of metaphor in thought. Thirdly and lastly, a resistance standpoint can address (a) metaphor serving as metaphor in communication between language users, the communicative dimension of metaphor.

Pragma-dialectics
Standpoints may not only vary according to their object of resistance but they may also differ according to the nature of the proposition they express. Pragma-dialectical argumentation theory distinguishes three types of propositions that can be expressed in standpoints: they can be prescriptive, evaluative or descriptive. Prescriptive standpoints recommend (or discourage) a certain course of action, evaluative standpoints express a judgement of facts or events, and descriptive standpoints describe facts or events (Van Eemeren et al., 2014). Different types of standpoints place different burdens of proof upon their protagonists. A common term in the domain of law, 'burden of proof' may often be associated with the legal duty that commands someone to provide factual evidence for the truth of their claim. Within the context of argumentation theory, however, the concept is defined in a broader way. Pragma-dialectical argumentation theory starts from the premise that argumentation is advanced in order to resolve a (potential) difference of opinion 5 ; through the exchange of arguments, discussants involved in the difference of opinion 3 Criticism that has been raised against DMT seems to stem from the misunderstandings that it aims to describe how metaphors are processed by actual language users (e.g. Xu et al., 2016;Gibbs and Chen, 2017). For a response to this criticism, please see . 4 Please note that in our discussion of the above examples we therefore purposely speak of a metaphor that "may often go unnoticed" and a metaphor that "may potentially influence language users' understanding", respectively. In relation to these specific examples we would like to stress that deliberateness is not something that can be determined by a factor such as word class (alone). A recent corpus study of 24,762 metaphors has demonstrated that adverbs, verbs, and prepositions are in fact used less frequently as potentially deliberate metaphors across different registers compared to nouns and adjectives in the same registers (Reijnierse et al., 2019). However, even though a metaphor's status as potentially deliberate or potentially non-deliberate may sometimes be related to the role of its word class and the register in which it is used, this status is directly dependent on whether the presence of the source domain in the referential meaning of an utterance can be observed in language use (also see Reijnierse et al., 2017). The latter can be established through DMIP, i.a. by establishing the presence of lexical signals and other contextual cues such as the use of direct metaphor and wordplay . 5 Pragma-dialectics recognizes that argumentative practices can vary widely depending on the communicative context in which they occur (e.g., Van Eemeren et al. (2014), pp. 557e563). There are argumentative activity types that are subject to explicit rules, such as legal trials or discussions in parliament; on the other end of the spectrum there are more informal settings in which argumentation occurs, like discussions among friends or in personal blogs, that can be characterized by more informal (social) conventions about the exchange of arguments. Within Pragma-dialectics, any context in which argumentation is advanced in order to convince anyone who (potentially) disagrees or doubts with a particular standpoint, is defined as a (potential) difference of opinion.
aim to test the acceptability of the standpoint at issue. This means that the protagonist of a standpoint is expected to support their standpoint with arguments when they are challenged to do so; if the protagonist fails to provide arguments for their standpoint, this renders the resolution of the difference of opinion impossible. The burden of proof is met if the antagonist is sufficiently convinced by the protagonist's arguments to withdraw their criticism or doubt (Van Eemeren et al., 2014). In this study we do not intend to take a normative perspective in the analysis of resistance standpoints. We will use the argumentation theoretical concept of burden of proof to better understand why language users put forward certain standpoints of resistance to violence metaphors for cancer to convince a (potential) antagonist of the acceptability of their standpoint. That is to say, as different types of standpoints place different burdens of proof upon their protagonists, these standpoints are typically supported in a particular way (Van Eemeren and Grootendorst, 1992). By taking this into account in our analysis of resistance standpoints, we can find out more about the protagonists' motivations for advancing their standpoint. The following examples may illustrate this.
If an arguer expresses a negative judgment on violence metaphors for cancer in the form of an evaluative resistance standpoint, the argument(s) they put forward in defence of the standpoint may be limited to their personal opinion regarding the metaphors in question. They may for instance argue that they dislike such metaphors because they do not like to be compared to army soldiers or because they prefer looking at their illness in a more peaceful, non-violent way. They could, however, also use factual evidence to support their evaluation. They might argue that they dislike violence metaphors for cancer because research has shown that these metaphors can lead to distorted conceptions of the disease.
Prescriptive standpoints are similar to evaluative standpoints in the sense that they can be based on personal opinion, factual proof or both. This likeness between evaluative and prescriptive standpoints is captured by Mohammed's (2016) remark that they are both value-based. Put differently, they concern two variants of value-based standpoints e "evaluative standpoints [are standpoints] in which a value-judgment about a certain state of affairs is expressed [and] prescriptive standpoints [are standpoints] in which a certain future action is recommended for its value" (Mohammed, 2016, p.3). 6 They differ, however, as to the argumentation that is typically provided in support of these two types of claims. For evaluative standpoints, arguments are aimed at convincing another party of the acceptability of the protagonist's judgment of a given state of affairs or to change one's beliefs about that state of affairs. For prescriptive standpoints, the protagonist's arguments are meant to change an antagonist's view of a given state of affairs ('the current situation is bad') as well as their behaviour ('one should (not) undertake this particular action in order to change this bad situation').
In other words, prescriptive claims carry more far-reaching implications than evaluative claims in that they do not only ask from an antagonist to accept the protagonist's evaluation of a particular situation, but also that a certain course of action needs (not) to be undertaken in order to change that situation. Following this difference in the argumentative support evaluative and prescriptive standpoints require, generally speaking, evaluative standpoint allocate a burden of proof on the protagonist that is lighter compared to a prescriptive standpoint.
As mentioned above, for evaluative and prescriptive standpoints arguments restricted to the protagonists' own subjective perspective suffice. This is different for descriptive standpoints: descriptive claims are typically supported in the form of facts or objective findings. One can argue that something is true (e.g., 'violence metaphors for cancer make people less likely to think of certain cancer prevention behaviours') based on another fact that has been proven to be true (e.g., 'research findings have shown that the use of violence metaphors in information messages about cancer does in fact reduce certain prevention intentions without increasing others'). Putting forward a standpoint on violence metaphors for cancer of the descriptive type thus imposes certain constraints upon the protagonist, who will base an argument in support of a descriptive standpoint upon objective findings rather than their subjective opinion.
The concept of burden of proof and the distinction between standpoint types are theoretical analytical instruments that can help us characterise the different (negative) positions language users express regarding the use of violence metaphors for cancer. 7 In actual discourse these positions may often resemble one another in different ways; the theoretical distinction between description, evaluation, and prescription can help pinpoint how they differ from one another when it comes to the main point of criticism a protagonist would like to convince their (potential) antagonists of. Our cases studies below will demonstrate how different types of resistance standpoints, placing different burdens of proof upon their protagonists, relate to the protagonists' motivations for advancing their standpoint. 6 I.e., prescriptive standpoints can recommend that one carries out a particular action or that one abstains from carrying out a particular action. 7 A body of recent studies within the pragma-dialectical framework has focused on the analysis of (various aspects of) argumentative activity types within the healthcare domain (e.g. Akkermans et al., 2018;Van Poppel, 2019). While taking into account the contextual factors that are of influence on the communicative e argumentative e phenomenon at hand, these studies starting from a pragma-dialectical perspective are specifically focused on argumentative characteristics of that phenomenon. As an anonymous reviewer rightly pointed out, the pragma-dialectical perspective taken in this study has as a consequence that other interesting information about resistance standpoints is left aside. However, we wish to stress that our study is focused on (demonstrating the added value of) getting an understanding of the different types of propositions that are put forward in resistance to violence metaphors for cancer; we thereby hope to contribute to the larger body of research that is being done on the (potential) shortcomings of violence metaphors for cancer. Several studies have previously reported other interesting findings on this subject, e.g. about the framing consequences of violence metaphors for cancer in relation to a patient's evaluations, emotions, and feelings of (dis)empowerment (Semino et al. (2016)).

Case studies
In this section, six examples of resistance standpoints will be discussed. Each of these examples demonstrates a different type of proposition in the standpoint (Subsection 3.1) or a different metaphor dimension as the object of resistance (Subsection 3.2). Three of the selected examples constitute relatively clear-cut cases of a specific standpoint type and another set of three examples concern illustrations of resistance standpoints addressing a particular dimension of metaphor. Subsection 3.3 will discuss the application of each of these two approaches to the other set of examples.
Two of the six examples that are discussed in this paper are from opinion articles published on websites aimed at providing information and support to patients with cancer, one example is from a cancer support website for peers and concerns the start of a discussion thread, and the remaining three examples are from blogs by cancer patients. All texts have been written by (former) cancer patients and are available on web pages that are publicly accessible. 8 The six examples have been selected from a larger corpus of instances of resistance to violence metaphors for cancer that was manually compiled by one of the authors. 9 Search engines on cancer support websites and social publishing platforms were used to search for discussions on violence metaphors for cancer that potentially contained expressions of resistance to these metaphors. A number of search terms were employed related to (1) the metaphors' target domain (e.g., 'cancer', 'cancer patient'), (2) the metaphors' source domain (e.g., 'fight' or 'warrior'), and (3) argumentative resistance (e.g., verbs like 'disagree' or 'dislike').
After a first categorization of the data into the relevant categories by one of the authors, all three authors analysed the instances of resistance individually. If each of the authors' interpretation of an example was the same, the analysis was regarded as intersubjectively reliable. If there turned out to be disagreement about a specific (aspect of) analysis, this was discussed among the three authors and a joint decision was made about the most likely interpretation(s) of the instance in question. In some cases, like for example (1) 10 , the discussion about different possible analyses of the same material led to new insights about this example's multi-interpretability. This was considered to enrich the outcomes of our analyses, sometimes bringing to the fore factors that complicate a straightforward categorization due to pragmatic aspects of language use.
The final selection of examples for discussion in this paper was based on the authors' judgment of an instance being suitable for illustrating the characteristics of a particular standpoint type or metaphor dimension in a resistance standpoint (Subsections 3.1 and 3.2, respectively) and, for at least two instances, for illustrating factors that may complicate a straightforward categorization (Subsection 3.3).

Evaluative, prescriptive and descriptive standpoints of resistance
Examples (1)e(3) below demonstrate how expressions of resistance standpoints can be differentiated according to the pragma-dialectical account of standpoint types.

Evaluative resistance standpoint
Example (1) is an excerpt from a blog post that contains an evaluative resistance standpoint. The blog post discusses the author's decision to quit chemotherapy. About this decision he writes the following: (1) On the "weenie" side, I guess this means I am not "fighting" cancer like a tough guy. To heck with that. That bout is rigged.
The author of (1) takes issue with a particular perspective on his illness that is based on metaphors of violence. He "guesses" that his decision to quit chemotherapy "means [he is] not 'fighting' cancer like a tough guy", a perspective he rejects by adding the conclusion "to heck with that". The standpoint concerns a value-based judgment on the part of the author: he negatively evaluates a specific way of looking at his illness experience, namely if it were conceptualised as a metaphorical fight. References to being "weenie" or "[not being a] tough guy" indicate the author's resistance to further associations based upon this mapping.
In the remainder of his blog post the author provides a number of arguments in support of his claim. He first discusses why he disagrees with the metaphor of 'fighting cancer' by arguing that this supposed fight is unfair. He then argues why he feels that the point of cancer treatment is not 'just' to prolong life e to him, living well has proven to be more important than living longer. Each of these arguments concerns further value judgments on the part of the author; by advancing his opinions on the 8 Using internet material for research purposes gives rise to a number of ethical considerations (Gustafsson et al. (2019). Yet "there is a lack of clear advice on ethical issues related to using internet material, such as blogs, as research data" (Gustafsson et al., 2019, p. 3) meaning that researchers using similar type of data follow different practices. In this paper we follow Demmen et al. (2015) and Seale et al. (2010) decision to consider internet materials that are publicly accessible as public data that does not require informed consent. We adhere to this view by understanding that it is based on the premise that the data is treated respectfully and used purposefully. For the current paper this means that we only present information about or parts of the data that we believe is necessary for increasing our understanding of standpoints of resistance against violence metaphors for cancer. In our case studies we give a direct quotation of an example's main standpoint (i.e., the exact words the author used to express their resistance standpoint on violence metaphors for cancer) and we paraphrase the argumentation that is given in support of this standpoint. In the paraphrases we have replaced specific references that may reveal information about the example's source by more generic references (following Gustafsson et al., 2019). 9 The corpus data is stored under permanent embargo in the UvA/AUAS figshare database at the University of Amsterdam. 10 This concerns the characterization of the metaphor dimension(s) addressed in example (1), which is discussed in Subsection 3.3. matter, the author seeks to meet the burden of proof the evaluative standpoint commits him to. This can be contrasted with the next example.

Prescriptive resistance standpoint
Example (2) is from a blog post by another cancer patient and demonstrates a prescriptive resistance standpoint. In this blog post the author argues why she prefers not to use violent language when she talks about her life with the disease. Moreover, she does not only discuss why she prefers not to use metaphors of violence such as 'warrior' or 'fighter'e she also means to discourage others not to use such language: (2) For crying out loud, don't call me a warrior.
In the remainder of her blog post the author motivates her resistance by mentioning that she does not feel empowered by picturing herself as a fighter unlike some other people do. She then provides arguments as to why she also explicitly refrains from using such metaphors herself. These arguments are specifically related to the author being a late-stage cancer patient who, according to the metaphorical comparison between cancer and a violent conflict, would be on the losing side of the battle. She subsequently extends the scope of her argument to include other cancer patients that are in a similar situation: according to the metaphors she contests, they should all be called failed warriors. Furthermore, being confronted with the gravity of her illness every day already, she does not need words of violence reminding her of this situation.
The author of (2) responds to the burden of proof of her prescriptive standpoint by arguing why she disagrees with violence metaphors for cancer being used to describe and talk to patients with cancer; in addition to expressing a negative evaluation of said metaphor use, the author of (2) provides arguments as to why she (thus) steers clear of violence metaphors for cancer. These can be read as arguments for her addressees to do the same e i.e., they simultaneously function as arguments to convince addressees not to use such metaphors when talking about cancer. This is different from example (1), in which the author argued his opinion without encouraging the reader to undertake a certain course of action, and can also be contrasted with example (3) below.

Descriptive resistance standpoint
The resistance towards violence metaphors expressed in example (3) concerns an example of a descriptive standpoint: According to the author of the blog post this excerpt was taken from, a problematic aspect to using metaphors like 'battle' and 'fight' in conversations about cancer is that it "puts the burden of healing on patients by turning them into winners and losers". In the remainder of her blog post the author mentions a few other examples of metaphorical expressions that are commonly used in discourse about cancer, including some specific implications that can be drawn from these metaphors. Calling cancer a 'battle', for instance, would imply that it must be bravely fought and that patients can win this battle as long as they fight hard enough. The author's standpoint in this matter is concerned with the problematic effects such implications may have on cancer patients. In order to support her standpoint the author refers to a blog post by another cancer patient in which exactly this issue is addressed. In this latter blog post it is discussed how it is not straightforward what 'beating' cancer means or what patients can or should be able to do to achieve this, yet talk about cancer as a 'battle' can evoke the feeling that one (should) have the ability to change the outcome of their disease. This, the patient argues, increases the heavy burden patients are faced with already in trying to overcome their disease.
Compared to the standpoints in (1) and (2), the standpoint in (3) is less of a subjective statement based upon the author's opinion. The author maintains that a given state of affairs is troublesome, the truth of which can be tested or proven in a relatively objective way. Put differently, some non-subjective standards can be employed to describe a situation as involving some problem or trouble. By expressing a descriptive resistance standpoint, the burden of proof the author of (3) takes upon herself involves argumentative support by means of a factual premise. The author meets this burden of proof for her descriptive standpoint by advancing an argument by example: the author cites a cancer patient who confirms to experience a burden of assumptions as a result of violence metaphors for cancer.

Resistance standpoints focused on a particular dimension of metaphor
Examples (4)e(6) below demonstrate how resistance standpoints can also be differentiated according to the dimension of (the contested) metaphor they address.

The conceptual dimension of metaphor
Example (4) is an example of a resistance standpoint that addresses the conceptual dimension of the violence metaphors for cancer contested by the author. The example is taken from a discussion thread on a cancer support website. The author, who initiates this thread, discusses how she struggles with conceptualising herself as a 'warrior' and her disease as a 'battle'. She writes the following: In the excerpt in (4), two words stand out in the light of our focus on metaphor dimensions as objects of resistance, i.e., "idea" and "see". They indicate the contested metaphors' role in the author's conceptualisation of cancer based on the metaphorical notion of the cancer patient as a 'warrior' and the illness experience as a 'battle'. One argument she raises in support of her resistance relates to the fact that these notions would imply that she is involved in a violent conflict with herself. This latter view also does not resonate with her preferred way of looking at the situation.
Both "idea" and "see" in excerpt (4) seem to point to the metaphors' manifestation at the level of thought. The fact that 'warrior' or 'battle' are metaphorical terms that are also used in talk about cancer and/or the fact that these metaphors play a particular communicative role in cancer conversations is not discussed in (4). This is not to say that the author may not object to these other dimensions of metaphor as well e in the expression of resistance in this particular excerpt, however, the standpoint addresses (metaphorical) thinking of a cancer patient as a 'warrior' in the first place. This can be contrasted with our next examples.

The linguistic dimension of metaphor
The resistance standpoint in example (5) below addresses the linguistic dimension of metaphor. The example is a short excerpt from an online opinion article written by a cancer patient. In the article the author discusses what she wishes people would stop saying to her in relation to her disease. She writes: (5) I wish people would stop using clich es [about cancer]. "You are so brave", "You are a warrior" […] The author subsequently argues why she feels that such 'clich es' 11 sound dismissive and why they do not seem to acknowledge the actual challenges cancer patients are faced with, especially when phrases such as "you are so brave" or "you are a warrior" are used to encourage patients to be positive: according to the author, patients should also be allowed not to remain positive at all times; moreover, believing that good things will happen or that a situation will get better does not cure patients of their cancer.
By explicitly expressing the wish no longer to be spoken to by means of the aforementioned 'clich es', the standpoint of the author is targeted at a particular manner of talking that is informed by the source domain of violence. According to DMT, this type of resistance would concern the linguistic dimension of metaphor: resistance is expressed against certain metaphorical expressions in language use e i.e., metaphorical expressions that (have come to) carry particular associations for the author that are not in line with her preferred way of looking at the target domain situation. In example (5) it specifically concerns violence metaphorical expressions that are used to convey allegedly misplaced messages of encouragement. Our next example will exemplify how resistance to this linguistic dimension of metaphor use can be distinguished from resistance to the communicative dimension of metaphor use.

The communicative dimension of metaphor
The resistance standpoint in example (6) constitutes an example of a resistance standpoint that addresses the communicative dimension of metaphor use. The example is taken from a blog post in which a cancer patient discusses her opinion on what are bad things to say to patients with cancer. In this post she argues why she thinks comparing cancer to a fight leads to a distorted understanding of the illness. Her standpoint about this matter reads as follows: She argues where she thinks the comparison between cancer and a fight goes wrong by pointing out the fact that some cancers are incurable and by noting that even patients who are diagnosed with a form of cancer that is potentially curable can die from their disease. These arguments are based on the conception that in an actual e non-metaphorical e fight those who fight hard (enough) are able to influence the outcome of the fight they are involved in. According to the author this does not apply to a 'fight' against cancer: it often does not matter how hard one tries to win as some factors are beyond a patient's control, whether it is the type of cancer one is diagnosed with, the stage at which the cancer is detected or another factor that affects one's chances of surviving the disease.
Unlike examples (4) and (5), example (6) explicitly addresses the mechanism that underlies metaphor use e the mechanism of (indirect) comparison. The author expresses resistance to what the metaphor does by relating one conceptual domain to another. Through the metaphor's cross-domain comparison, concepts from the source domain are transferred onto the target domain. In the particular example of (6) the idea that a fight will be won by the party who fights hard enough is deemed problematic when it is applied to the metaphorical scenario of the cancer patient 'fighting' against their disease. Here lies a subtle difference with example (5): in (5) the resistance seems principally aimed at violence metaphorical phrases that are meant to encourage a patient or to lift their spirits, something which the author judges to be admonishing e it is the tone of these phrases or the emotion or attitude they are understood to express that the author appears to resist in the first place. The author of (6) places more emphasis on the comparison underlying the metaphor use, transferring implications of a literal fight onto the metaphorical 'fight' against cancer, which she considers to be inapt.
Example (4) read: "I don't like the idea of being a Warrior. I don't see cancer as a battle, because I don't want to be at war with my self [sic]." In this standpoint the author expresses a negative judgement of the warrior metaphor, which is indicated most clearly by the first verb used in the example: the author "[does not] like the idea of being a Warrior". Her arguments for this standpoint have to do with her observation that she also does not want to be 'at war with herself', which would be the underlying implication of picturing herself as a warrior, or her disease as a battle. These are subjective evaluations on the part of the author. Something similar can be observed in (6): The standpoint in (6) reads that the author "can't stand it when people compare cancer to a fight" (italics added). Here too the use of the verb "can't stand" points to the author's subjective evaluation of the object of her resistance, i.e., "people comparing cancer to a fight".
Compared to (4) and (6), the identification of the type of standpoint in example (5) requires some more careful consideration. Like in (4) and (6), the finite verb in the standpoint of (5) provides a starting point for the determination of the standpoint type. The author of (5) writes she "[wishes] people would stop using clich es [like] 'You are so brave', 'You are a warrior' […]" (italics added). While her wish can be understood to concern a value-based statement about a certain state of affairs, it may in fact communicate something else. The author's expression of a wish for a situation to change can be interpreted as a request: the people referred to in the standpoint are requested in an indirect way to change their behaviour, or to undertake the action the author argues to be desirable. The example can therefore be reconstructed as an implicit standpoint 12 of the prescriptive type: the author does not explicitly ask for people to change their behaviour, but by formulating the wish for this to happen the author proposes a prescriptive claim in an implicit manner.
We will now turn to examples (1), (2), and (3). The standpoint in (1) read: "On the "weenie" side, I guess [my decision to quit chemotherapy] means I am not "fighting" cancer like a tough guy. To heck with that." The quotation marks used by the author can be read as markers of irony meant to accentuate that the author disagrees with the metaphor. This interpretation may be supported by the fact that quotation marks are also used for "weenie". The marks subvert the actual meaning of the words in question: the author's decision to quit chemotherapy is by no means a sign of cowardice, just like the notion that one can 'fight' cancer does not compare with the actual experiences cancer patients go through. This interpretation might indicate that the conceptual dimension of metaphor plays a main role in the author's resistance: he argues against conventional thinking that cancer can or should be 'fought' against.
Arguably, the quotation marks in (1) may also concern a reference to the use of the expression of 'fighting cancer' in conversations about cancer, emphasizing that this is how people talk about the disease. From the wording of the standpoint it seems difficult to determine what possible role is played by the metaphor's source domain as a distinct referent in the meaning of the contested metaphorical expression e in other words, whether the resistance is also targeted at the metaphor's use as metaphor in communication (i.e., DMT's communicative dimension). The standpoint does not give us any particular clues about this, but if we look at the arguments that are provided in support of the standpoint in (1), we see that the author does criticize the cross-domain comparison underlying the metaphor use. All of the above considered it thus seems more difficult to see one particular metaphor dimension standing out as the main target of resistance in example (1) compared to (4), (5) and (6) discussed in Subsection 3.2.
The standpoint in (2) read: "For crying out loud, don't call me a warrior". This standpoint may also be related to the linguistic as well as the communicative dimension of metaphor: the author asks her readers not to use certain violence metaphorical expressions when talking about cancer (e.g., 'don't call me …'). Like in (1), from the formulation of the standpoint alone it cannot be told which possible role is played by the communicative dimension of the contested metaphors e for instance, the verb 'call' does not point as clearly to a relation to the use of metaphor as metaphor in communication like the verbs 'compare' (cf. example (6)), 'liken' or 'parallel' do. The arguments that are provided in support of the standpoint, however, seem to be based on an extension of the contested metaphor: the author takes the cross-domain comparison between patients and warriors to mean that she and other late-stage cancer patients should be called 'failed warriors'. These arguments can be interpreted to address the inappropriateness of comparing patients to 'warriors', which indicates a possible role of the communicative dimension in the standpoint of resistance.
The standpoint in (3) read that "[the] trouble with using [metaphors such as 'battle' and 'fight'] to describe cancer is [that] it puts the burden of healing on patients by turning them into winners and losers". The author speaks about metaphors that are used to 'describe' cancer which can again be related to the linguistic as well as the communicative dimension of metaphor use. Moreover, the second part of the standpoint, about patients experiencing a burden of healing, can be connected to the role of metaphor in thought. For example (3) it thus also seems more complex to pinpoint one particular metaphor dimension that stands out clearly as the main target of resistance.
In sum, our analyses in Subsections 3.1 and 3.2 demonstrated relatively straightforward examples of a particular metaphor dimension or standpoint type. Each of the examples contained certain clues or keywords that made it possible to connect a particular expression of resistance to one standpoint type or metaphor dimension. The above analyses show how it can sometimes be more difficult to determine which dimension of metaphor forms the main focus of resistance, or which 12 For the pragma-dialectical perspective on implicit or unexpressed premises, see Gerritsen (2001). standpoint type is used; furthermore, for some expressions of resistance the (con)text lacks clear indications for the analyst to determine if one e and which e dimension of metaphor use plays a role in the emergence of the arguer's resistance.

Conclusion
The first objective of this paper was to give an insight into the different types of resistance standpoints that are expressed in critical responses to violence metaphors for cancer as well as the different metaphor dimensions these standpoints are focused on. A second aim of this paper was to demonstrate the value of a close analysis of standpoints of resistance to metaphor from a combined argumentation theoretical and metaphor theoretical perspective.
In line with pragma-dialectical argumentation theory, three types of resistance standpoints against violence-metaphors for cancer were defined: evaluative resistance standpoints expressing a negative judgement of violence metaphors for cancer; descriptive resistance standpoints describing factual observations concerning the contested metaphor use; and prescriptive standpoints discouraging the use of violence metaphors in relation to cancer. Our case studies in Subsection 3.1 exemplified the differences between these different types of resistance standpoints. The analyses also demonstrated how each standpoint type has consequences for the protagonists' burden of proof. As a result, a more detailed picture could be formed of the protagonists' motivations for advancing their standpoint or their aims to convince (potential) antagonists of the validity of their opinion, the truth of a fact, or the necessity to undertake a certain course of action.
Deliberate Metaphor Theory guided our analysis of the propositional content of resistance standpoints. More specifically, we categorized resistance standpoints according to whether they seemed to be principally related to a) conceptualising cancer in terms of the source domain of violence (DMT's conceptual dimension of metaphor), b) talking about cancer by means of violence metaphorical expressions independent of the underlying cross-domain comparison between source and target domains (DMT's linguistic dimension of metaphor), or c) communicating about cancer in such a way that the source domain of violence functions as a distinct referent in the meaning of a metaphorical utterance (DMT's communicative dimension of metaphor). This concerned a novel application of DMT. Our case studies in Subsection 3.2 showed that in some expressions of resistance particular dimensions of metaphor use did seem to stand out as the main targets of resistance: certain keywords appeared to indicate that some expressions of resistance were focused on one specific dimension of metaphor use more than another.
The case studies in Subsections 3.1 and 3.2 presented relatively clear-cut examples of a particular standpoint type or metaphor dimension. In Subsection 3.3 we also discussed less-straightforward cases in which standpoints displayed characteristics of more than one standpoint type or metaphor dimension. In addition, it was shown that sometimes the (con)text of an example provides too little information for the analyst to make a sound judgment about which dimension of metaphor potentially formed the primary target of resistance.
In sum, the analyses in this paper proposed new insights about how metaphors are perceived and which specific aspects of their use are deemed inapt or inappropriate by actual language users through a close analysis of resistance standpoints. The paper thereby aims to contribute to the larger body of research that is done on the (potential) shortcomings of violence metaphors for cancer. The paper has also argued that a combined metaphor and argumentation theoretical approach to the analysis of resistance standpoints renders a more detailed account of resistance to violence metaphors for cancer compared to an exclusively argumentation theoretical or exclusively metaphor theoretical approach.
Lastly, the cases analysed in this paper may be seen as illustrative examples of how the combined metaphor and argumentation approach can potentially be applied more generally, also for other metaphors or in other contexts of metaphor use. For each instance of resistance to metaphor, regardless its source or target domain, a close analysis from the perspectives of Pragma-dialectics and DMT can improve our understanding of the negative points of view language users hold towards the metaphor in question.