What Am I, a Piece of Meat? Synecdochical Utterances Targeting Women

In a September 2004 interview, Donald Trump agreed with Howard Stern’s statement that his daughter Ivanka is “a piece of ass.” This utterance is a synecdochical utterance targeting women (SUTW), by which I mean that its form is such that a term for an anatomical part is predicated of, or could be used by a speaker to refer to, a woman. I propound a theory of what SUTW speakers do in undertaking an SUTW on which the SUTW speaker prompts the hearer to engage in a certain derogatory pattern of associational thinking — that is, taking a “perspective” in Elisabeth Camp’s sense— on the female subject. This perspective is one that reduces her to the bodily part in question — that is, fragments her (reduces her to a part) and biologizes her (characterizes her as mere living tissue). Essentially, the hearer thinks of the woman as a “piece of meat . ”


Introduction 1
In a September 2004 interview with Howard Stern, Donald Trump agreed with Stern's description of his daughter Ivanka: "[She's] . . . a piece of ass" (Kaczynski, Massie, and McDermott 2016). Stern and Trump's utterance is a synecdochical utterance targeting women. By a "synecdochical" utterance, I mean an utterance of a sentence consisting of a part term and the predicate "is a [bodily part]." 2 Synecdochical utterances targeting women (henceforth, SUTWs) are ones in which the "whole" is a woman and the "part" a bodily part, and in which that part is associated with female bodies. 3 Stern and Trump's utterance is clearly derogatory. My central question is, what kind of derogatory act does an SUTW speaker perform?
I offer a theory on which the SUTW speaker prompts the hearer to engage in a certain derogatory pattern of associational thinking-that is, taking a "perspective" in Elisabeth Camp's sense-on the female subject. This perspective is one that reduces her to the bodily part in question-that is, fragments her (reduces her to a part) and biologizes her (characterizes her as mere living tissue). Essentially, the hearer thinks of the woman as a "piece of meat." 4 My theory explains not only derogatoriness but also other data that show that the speaker engages in an act other than conveying the literal meaning of the uttered sentence: the inability of a speaker to deny that undertaking this other speech act ( §3.1) and her choice of a bodily part rather than a salient alternative ( §3.3). 5 I can explain why hearing an SUTW offends us ( §3.2), a feature that supports thinking of this other act as derogatory. 6 The offense traces back to the perspective the speaker prompts the hearer to take on the female subject. The perspectival theory explains data pertaining to which kind of act this derogatory act is-and data that pull us in diametrically different directions, as well ( §4). This includes data that speak in favor of a view on which the additional act that a speaker performs is conveying cognitive content: some paraphrases seem apt, and a conversational participant can refer back to or (dis)agree with that content. Other data steer us toward a noncognitive analysis: paraphrases do not seem to fully capture what is conveyed, what seems apt depends 3 For an utterance to constitute a SUTW, it must also be used in a genuinely synecdochical way. See §2. 4 My point in this paper is to argue for what constitutes an SUTW's content, that an SUTW speaker evokes a perspective, rather than to argue for the mechanism by which the content is conveyed. However, my position is that an SUTW is a perlocutionary act, whereby the speaker causes the hearer to take on a perspective. The hearer's perspective-taking is a perlocutionary effect. 5 One might object that I preclude the possibility that the SUTW speaker conveys synecdochical content as a matter of the content the sentence literally expresses, as Josef Stern (2000) argues for metaphor. There have been numerous criticisms of such an account. Camp (2005), for example, provides several. For this reason, I limit my inquiry to accounts on which SUTW content is conveyed via nonliteral means. 6 Offense is distinct from derogatoriness (Hom 2008). Derogatoriness is a feature of the utterance, whereas offense is a psychological effect we experience when we hear a derogatory utterance. These can easily come apart as when a derogatory utterance fails to cause one offense because she has become desensitized to it. on context, the speaker can coherently deny any paraphrase, and her ability to plausibly deny any paraphrase varies by context. My project has especial importance for feminist philosophy. Any of the associations which an SUTW evokes are familiar from the literature on objectification: that a woman lacks agency, that it is permissible to exploit her for her sexual and reproductive capacities, that parts of her body are interchangeable with another woman's, and so forth. These are traits that Martha Nussbaum (1995) and Sandra Bartky (1982, 129-130) recognize as characteristic of objectification. Rae Langton (2009) identifies reduction to body, of which an SUTW is a clear linguistic manifestation, as a feature of objectification. This suggests that characterizing women as meat is a way of objectifying women. SUTWs, by inviting the hearer to think of the woman in a way that violates her autonomy, might also be expressions of misogyny on Kate Manne's (2018, 84-86) feminist analysis. If my theory is correct, then one can objectify a woman or undertake a distinctively misogynistic act using speech. This is the plan. In section 2, I elaborate on the criteria for an SUTW. I expound on the reasons for why we should think that an SUTW speaker does something other than convey the SUTW's literal content and something derogatory in section 3. Then, I enumerate features of the kind of content an SUTW speaker conveys in undertaking this derogatory act ( §4). Next, I present a theory of what the speaker does-namely, causing the hearer to take a perspective on the woman that reduces the female subject to a piece of meat ( §5)-and make clear how it fits the characteristics of an SUTW to be explained ( §6). Lastly ( §7), I explain the importance of my theory for socially engaged philosophy of language.

What Are Synecdochical Utterances Targeting Women?
We can call Stern and Trump's utterance above, (1). Utterance (2), which is taken from a blog post about Eurovision star Gaitana, is another instance of an SUTW.
(2) She's a pair of legs (W. Adams 2012). 7 Linguist Zoltán Kövecses (2006, 160-161) also notes that "leg," and "ankle" for that matter, are terms used to refer to women. We are familiar, too, with expressions like "showing some leg," which does not make reference to a particular woman.
Utterance (3) was commonly heard on social media during the 2016 American presidential campaign. Given that the part term in (3) is so incendiary, I use the circumlocution c-word and will do so whenever I refer to the term.
We should now identify those aspects of SUTWs that make them SUTWs. Firstly, SUTWs typically exemplify some variant of a simple, representative form: "pair of legs" is in the predicate position, and "she," the term used to refer to the woman, is in the subject position. 8 In contrast, (4), taken from a ZZ Top song, lacks the representative synecdochical form.
The speaker of (4) does not say that the woman "is" a pair of legs, but that she "got" or possesses legs. 9 A condition on form, though, is not the only requirement for a SUTW. The utterance must be genuinely synecdochical, as well. To be genuinely synecdochical the bodily part term must (a) express the sense of the relevant bodily part. I use braces to mark the sense of the term in what follows. In (1) "piece of ass" expresses {posterior of the pelvic area}. For (2), the sense "pair of legs" expresses is {pair of legs}, the limbs people typically use to walk. The next condition concerns the use of "is." In a genuinely synecdochical utterance, the "is" is not the "is" of attribution: the speaker does not impute the part to the whole as one does a property. Rather, she represents the whole as a part and so, the "is" in a genuinely synecdochical utterance is the "is" of identity.
Apart from constituting a genuinely synecdochical utterance, there is another condition for an utterance to be an SUTW: (b) a constraint on the type of bodily part. The part must be characteristic of female bodies. 10 8 I say "characteristic" since there are other forms SUTWs can take as in "She and her legs: they're one and the same." These variants are all ones in which language other than a form of the verb "be" is used to express identity between the female subject and the bodily part term. 9 Although (4) is not an SUTW, it is offensive in many contexts. See §6 for discussion on (4). 10 I say "female bodies" rather than "women" for "female" names a sex, whereas "women" refers to a gender. Bodies are "female" by virtue of biological features such as hormones, chromosomes, and anatomical parts whereas gender is convincingly argued to be socially constructed (see, e.g., Haslanger 2012). One need not have a body sexed as "female," though, to be the subject of an SUTW. Anyone perceived as "female" regardless of that person's gender identification (or lack thereof) can be the subject of an SUTW.
The condition that the part should be characteristic of female bodies does not imply that the part is unique to women. The predication of "pair of legs" to a woman is an SUTW-but men have legs, too. It also does not imply that every bodily part that a woman has qualifies as the right sort of part. Earlobes, for example, are not distinctive of women so predicating "ear lobes" of a woman (suppose one known for her appearances in earring advertisements), as the speaker of (5) does, is not undertaking an SUTW. 11 (5) She's a pair of earlobes.
Types of bodily parts that spring to mind when we think of those characteristic of female bodies are those directly involved in reproduction and childbearing, like a uterus. The SUTW (6) is an example.
(6) You [to a female person] are only a uterus (Eichler 2013). 12 Secondary sexual characteristics like breasts, waists, or bottoms are also often predicated of women-as we see in (1). Legs are also commonly used to represent women. Consider, for instance, a familiar representation of a woman drawn from film: a woman putting on her stockings.
An SUTW, then, is an utterance (i) whose subject is a woman, (ii) in which the bodily part predicated of the subject is characteristic of female bodies, and (iii) that is used in a genuinely synecdochical way-that is, in which the part term in the predicate expresses the sense of the relevant bodily part and the "is" connecting the female subject and the bodily part is the "is" of identity. 13 11 It is possible that in a society in which a woman's earlobes are associated with female bodies (for example, by being sexualized) that (5) could be an SUTW. That is, which parts are sexualized is culturally contingent. I am indebted to an anonymous referee for pressing me to make this explicit. 12 In a similar vein, Mary Beth Whitehead, the defendant in a landmark case on the custody rights of surrogate mothers, said of the couple that contracted her to bear their child that they viewed her as "a uterus with legs" (Hanley 1987). Concerning recent attempts to ban abortion in many Southern states, medical doctor Erin King said in a personal email to the author (June 4th 2019), "It is fair to say that the uterus is the most regulated thing in the state of Missouri." 13 It might seem that parts "characteristic of female bodies" are simply sexualized parts. Yet, not all parts characteristic of women like a uterus (as in (6)) are sexualized. My thanks to an anonymous referee for asking me to address this point.

Synecdochical Utterances Constitute Derogatory Acts
Now that we have a grasp of what an SUTW is, one might wonder the following: why should we think that the speaker does anything other than convey the SUTW's literal content, and why a derogatory act? We can refer to this nonliteral act as the "derogatory" act. This alternative language enables us to reformulate the question: why should we think that the speaker performs the derogatory act? I provide three reasons: illocutionary noncancelability ( §3.1), the psychological effects of offense and reduction an SUTW triggers ( §3.2), and the speaker's choice of a bodily part term over relevant alternatives ( §3.3).

Illocutionary Noncancelability
To see the first distinctive characteristic of SUTWs, reconsider (2). As a genuinely synecdochical utterance, the speaker asserts the face-value significance of the utterance, "She's a pair of legs." The part term expresses {pair of legs}.
However, this cannot be all the speaker intends to do by asserting this, for the assertion that a woman is identical with the relevant bodily part is patently false. 14 Rather, the speaker must intend to do something else using her utterance, on pain of failing to be a cooperative conversational participant. For this reason, the speaker of (2) cannot be merely the asserting that "she is a pair of legs." The speaker must be doing something in addition to conveying the utterance's face-value significance, "She's a pair of legs." The point is that this cannot be the only thing she intends to assert.
We can call this characteristic of SUTWs illocutionary noncancelability. Illocutionary noncancelability is a plausible reason to think that the SUTW speaker does something other than convey the literal content of the sentence that she utters.

Psychological Effects: Offense and Reduction
Moreover, targets and empathetic hearers experience a certain kind of psychological effect in understanding what an SUTW speaker conveys: offense (Hom 2008). Some women describe the type of psychological effect they experience more specifically, whether as targets or empathetic hearers, as the feeling that the woman is most important in virtue of the part. I call this effect reduction. One testimonial is Glennon Doyle's (2016): "Misogynists call women cunts when they are trying to . . . remind us that what our identity boils down to is: our [c-word]." Aside from testimonials like Doyle's, the sorts of responses some women give to these remarks evince that they make them feel reduced. The form of reply is to insist that the woman is more than the part with which she is identified. One woman says that she is "more than a uterus" (Eichler 2013;loribeth 2013). Political journalist and author Amanda Marcotte (2016) responded to a remark made by Killer Mike about Hillary Clinton that "Clinton's remarkable career isn't enough to prove that women can be worth more than a 'uterus.'" Not all targets of SUTWs find SUTWs to be reductive or offensive. Bethenny Frankel (2017) captioned a photograph of herself that she posted on Instagram, in which the relevant anatomical part figured prominently: "A lotta people refer to me as an ass." Contextual information-most saliently, knowledge about Frankel's selfobjectifying attitude towards her body-make it clear that she perceived her admirers' synecdochical characterization as flattering. Of course, the fact that Frankel in this case is not in fact offended, or does not feel reduced, by her characterization as a bodily part does not entail that a negative reaction to the utterance is not warranted (Bolinger 2017).

Choice of Bodily Part Term
Another datum that needs to be explained is the following: why does the speaker predicate a part term, and a bodily part term, as opposed to some other language? On the assumption that the speaker is a cooperative conversational participant, there must be a reason why she chose to predicate a term that is both bodily and a part of the female subject. More specifically, using both a term that is bodily and a part term must enable the speaker to perform the act that she intends to undertake. 15 In order to see that the speaker does make such a choice, it would be helpful to consider what the alternative language would be. One point of contrast is with rhetorically similar nonpart terms. Why did Stern and Trump in (1) predicate "piece of ass" of Ivanka rather than nonpart terms, which are also not biological in nature, such as "fox," "babe" or "hot young thing"? Or, for (6), why select the bodily part term "uterus" rather than one that is not a part but is biological or evokes animals, like "broodmare"? Indeed, why select a bodily part term rather than a term that is nonbodily but, depending on your conception of the relevant whole, could be a part, such as "skirt"?
On the assumption that the speaker is cooperative, we should infer that the speaker's choice of a term that is both bodily and a part plays some distinctive role in what the speaker does in uttering an SUTW.

What Kind of Derogatory Act Does a Synecdochical Utterance Constitute?
In the previous section, I provided three reasons for why an SUTW speaker performs what I called "the derogatory act"-that is, does something other than convey the literal meaning of the SUTW and something derogatory. Now we should consider data pertaining to what kind of act this is.

Data That Pull Us in a Cognitive Direction: Aptness, Reference, and (Dis)agreement
The first important feature of SUTWs to notice is that we can provide reasonable paraphrases of what the speaker conveys in performing the derogatory act. That is, some articulations of what the speaker conveys are apt. Take (1). Paraphrases (1.1P)-(1.3P) are apt specifications of that Stern and Trump convey.
(1.1P) Ivanka is just valuable to gawk at and use as a sex object.
(1.2P) Ivanka isn't intelligent and lacks the ability to determine her own life.
(1.3P) Ivanka is a piece of meat.
Our ability to paraphrase what the speaker conveys might lead us to think that the derogatory act an SUTW undertakes is conveying cognitive content. Two other features of SUTWs seem to support this hypothesis. Camp (2007) identified both of these as characteristics of metaphor and argued that they speak in favor of this kind of analysis. The first is that a party to the conversation in which (1) is uttered can refer to a paraphrase like (1.1P) later on. The other is that a speaker can agree or disagree with paraphrases. Suppose that the speaker of (1) and his hearer, both of whom are struggling actors, are discussing the ability of a female acquaintance of theirs to obtain a coveted role. The hearer could agree with what the speaker says using (1.1A)-(1.3A).
(1.1A) Yeah, she's just something to look at, dress up, and parade on the set like a doll.
(1.2A) Yeah, she doesn't have much of a mind of her own.
(1.3A) Yeah, she's just a piece of meat.
The hearer can also disagree with the speaker by insisting on the target's identity as a person-and so meriting the moral status appropriate for a person (as in 1.1D or 1.3D)-or insisting on her possession of recognizably human features like intelligence, autonomy, and other character traits like diligence, (here, acting) talent, and creativity (1.2D).
(1.1D) No way: she's intelligent, hardworking, and genuinely talented! (1.2D) No way: she's been a major creative force behind this project! (1.3D) She's not meat-she's a person and deserves to be treated like one! These three features-the aptness of some paraphrases of what the speaker conveys and a conversational participant's ability to refer to what is conveyed or (dis)agree with it-give us reason to think that the speaker conveys something cognitive, particularly propositional, in nature. The next set of data leads us to a different conclusion that the speaker conveys something noncognitive.

Data That Pull Us in a Noncognitive Direction: Ineffability, Deniability Patterns, and Contextual Variability
One feature that militates against a cognitive analysis is an SUTW's coherent deniability pattern. By this, I mean the speaker's ability to coherently-however disingenuously-deny paraphrases of what she conveys in performing the derogatory act. Notice that coherent deniability is not the same as plausible deniability. An objector can still find it plausible (even believe) that the speaker conveyed the specification of the content attributed to her, even though the speaker's denial of that explication is compatible with what is said. With this distinction in mind, Stern and Trump could coherently deny that any of (1.1)-(1.3) is an accurate paraphrase of what their utterance conveyed.
Indeed, Stern and Trump could coherently deny any particular paraphrasenot just those expressed in (1.1)-(1.3). If we replaced the explication with ones as varied as "she's something that one can use as one pleases," "she's disposable," or "she's a piece of meat interchangeable with any other," Stern and Trump could deny conveying any of these. The ability of an SUTW speaker to coherently deny any paraphrase is some evidence to think that the SUTW speaker does not convey cognitive content at all.
An additional reason to doubt that a speaker conveys cognitive content is that the aptness of the content attributed to the speaker is not equally apt in every context. The explication of (1) as specified in objection (1.1P) would become more apt than, say, (1.2P) if (1) is uttered about a woman in a context in which the speaker's sexual interest in the referent of "she" is apparent. To be clear, in such a context, the aptness is comparative: (1.1P) seems more apt than (1.2P), but (1.2P) would still be a reasonable paraphrase. However, if (1) were uttered in a setting that the speaker referred to earlier in the conversation as a "meat market" (e.g., a gym) or in a context in which women are treated like "meat," the paraphrase offered in (1.3P) becomes more apt than the alternatives. There is testimonial evidence that pornography is a context like the latter. An anonymous actress says of her experience that pornographers "treat us like meat" (as quoted in C. Adams 2015, 72). Linda Lovelace, a well-known former adult-film actress, describes her experience auditioning for a director as follows: "[He] looked me over like a butcher inspecting a side of beef" (72).
Moreover, the speaker's ability to plausibly deny conveying some content also varies by context. It is harder for the speaker to plausibly deny (1.1P) if the utterance is made in a context in which the speaker's attraction to the female subject is known to the objector. The speaker can less plausibly deny conveying (1.3P) in a setting as degrading as Lovelace's audition. To be clear: the speaker could still coherently deny conveying (1.1P)-(1.3P), but the hearer is less likely to accept the sincerity of the speaker's disavowal.
One might object that a speaker's ability to coherently deny any particular paraphrase, the contextual sensitivity of a paraphrase's aptness, and plausibility deniability might still be compatible with an analysis on which a speaker conveys indeterminate and contextually sensitive cognitive content. Even if this is true, another feature more strongly recommends a theory on which the content is noncognitive. Although paraphrases (1.1P)-(1.3P) are apt, we also detect that any one of them-or even all of them collectively-do not exhaust what the speaker conveys.
That is, what the speaker conveys is, to some extent, "ineffable" (Potts 2007, 176;Blakemore 2011). The inability to fully articulate the content is some reason to think that there is no articulable content and-since cognitive content is characteristically articulable-no cognitive content (Potts 2007, 177).
We need a theory that can accommodate the data in sections 4.1 and 4.2, which move us in divergent directions. On the theory that I present in the next section, an SUTW speaker conveys indeterminate, contextually sensitive content-but content that by its nature has both cognitive and noncognitive elements.

What Synecdochical Speakers Targeting Women Do
The sort of act that an SUTW speaker undertakes exploits both features of the SUTW's sentence structure and elements of its context of utterance, especially attitudes towards women in the broader social context. First, we should be clear about the linguistic features of an SUTW enabling a speaker to perform the derogatory act. An SUTW, as a genuinely synecdochical utterance, has a representative sentence structure: "is" links the term referring to the female subject and the bodily part terms, and the sense of "is" is the "is" of identity. The bodily part term expresses the sense of the relevant bodily part. Hence, the sentence (2) literally expresses that the referent of "she" is identical with her legs. 16 My claim is that this kind of sentence is distinctively well suited to be used to convey the content that accounts for the data observed above. In hearing a sentence that literally expresses that a female subject is identical with a bodily part characteristic of female bodies, the hearer undergoes an associational process. This process is not a mere effect but also enables the hearer to interpret what the speaker conveys.
In particular, I claim that the speaker undergoes a particular interpretive process whereby she thinks of the female subject in a way that reduces her to a bodily part (informally, as a "piece of meat"). This is to think of her in a way that is both fragmenting and biologizing. By being fragmented, the woman is characterized in terms of a part of herself. In being "biologized," I mean that the woman is thought of in a way that reduces her to a purely biological entity. She is a mere living-sentient but mindless-tissue. To think of a woman in a "biologizing" way is to think of her in the way people tend to (inaccurately) think of animals: as brutes lacking intellect. We can also specify my claim in terms of the utterance: an SUTW is associated with a way of thinking as reduced to a bodily part. In order to fully understand what I have in mind we need to become clear about what is meant by thinking as-namely, taking on a perspective. What makes the perspective one that reduces her to a bodily part is explained in terms of taking on a perspective with its particular contents and the affective responses it inspires. Once I have explained the elements of the perspectival theory, I make explicit how it explains the data in section 6.
By "think [of a female subject] as," I mean take on a perspective in Elisabeth Camp's sense (e.g., 2013, 2017). A perspective on, say, rats is an interpretive tendency that we have "to notice, relate, and respond in certain ways" to rats. For instance, if I encounter a rat, I am apt to notice her long tail and dark brown fur, become fearful, and avoid her. Its constituent components vary in their natures. A perspective might include concepts, encoded as descriptive properties (e.g., red, rectangular), evaluatively laden "thick" properties (magnanimous, obstreperous), or "thin" properties (superlative, awful). Despite the fact that they might include cognitive elements like concepts, perspectives are nonpropositional. The inclusion of thick and evaluative thin terms in a perspective is also one way in which a perspective has affective elements.
There are two important structural relations among the properties comprising a perspective. Properties that are central "explain, cause, or otherwise motivate" the other contents (Camp 2017, 50-51). My perspective on rats contains properties like [repulsive] and [untrustworthy]. The property [sneaky] is more central: it causes and explains features like [untrustworthy]. Taking a perspective on someone frames how I interpret her behavior and appearance: when I take a perspective on my colleague as a rat, I am disposed to see the way she moves her nose as the way that a rat twitches her nose or to hear her voice as squeaky. This leads us to a second way in which perspective is partly affective. Certain attitudes and emotional responses are also constitutive of perspective-taking. In seeing someone as a rat, I feel suspicion and-most appropriate to a rodent-disgust.
Some features are also more prominent than others (Camp 2017, 51). Features prominent in my perspective on rats like [squeaky] and [bewhiskered] enable me to identify which features matter in determining which things are rats. Notice that prominent features are not ones that must be possessed. Rats need not have whiskers to be rats. Nonetheless, whiskers help me identify rats.
The components constituting the characterization are structured into a coherent whole, in a way that is intuitively "fitting" (51). The sense of "fitting" cannot be articulated in terms of any explicit principle: it's the sort of norm one knows something conforms to intuitively. In this way, "fittingness" is like an aesthetic norm. That is, these contents "fit" together in the way that, say, a chair with a simple, sleek design fits into a room otherwise furnished in a modern style.
Camp identifies slurring utterances (2013) and metaphorical utterances (2017) as utterances associated with perspectives. For the theory of SUTWs that I am presenting, the particular process evoking by hearing a metaphorical utterance is pertinent: restructuring her perspective on the female subject "in light of" the part predicated of the female subject. There are multiple elements here: the perspective on the female subject as a woman (Ivanka in (1)), the perspective on the bodily part ("piece of ass"), and the process of restructuring the former using the latter.
Let us consider how this process using a metaphorical utterance like "Peter Pettigrew is a rat" is associated with two perspectives: one on Pettigrew and one on rats. Pettigrew was a character from J. K. Rowling's Harry Potter series known for his treachery. In hearing this utterance, the hearer restructures her perspective on Pettigrew using her perspective on rats. "Restructuring" is the process by which the hearer matches those features shared in common between these perspectives and then alters the configuration of the components in the perspective on Pettigrew based on the matches. For instance, this alteration could consist in features like [sneaky] or [untrustworthy] becoming more central in one's perspective on Pettigrew. Thus, when I hear this utterance, I think of Pettigrew as having these traits and in a way that these traits intuitively suit him. Remember that there is an affective component to perspective-taking: in taking on a perspective, I not only think of Pettigrew as having traits fitting for a rat but also feel towards him as I would a rat.
On my view, we should understand what an SUTW speaker does as thinking of the female subject in a way that fragments her (thinking of her as a part) and biologizes her (thinking of her as exclusively bodily in nature, as opposed to intellectual or spiritual). 17 I am also claiming that perspectives effectively model this way of thinking: we should understand what it means to think of the woman in a way that reduces her to a bodily part as undertaking a perspective on the relevant woman (and more specifically, one shaped by a perspective on the bodily part).
Analogous to the structure of thought associated with the rat utterance, the hearer uses a perspective on the bodily part to frame how she thinks about the woman of whom that part is predicated. So, what is the perspective of some bodily part like? A perspective on a bodily part reflects our understanding of the nature of an anatomical part, containing elements like [animal], [purely biological], or [exclusively bodily], as well as ones associated with paradigmatically biological entities: nonhuman animals, particularly mammals. (By "purely" or "exclusively" biological or bodily, I mean containing no elements pertaining to intellect, subjectivity, or spirituality.) Thinking of an entity as an animal-on our perspectival model, connecting the concept of that entity (say, [women]) with [animal]-is tied to thinking of it as lacking traits thought to be unique to human beings (Harris and Fiske 2006;Haslam 2006). One of the most prominent of these traits is rationality (Haslam 2006). Thus, by associating something with animals, one associates it with irrationality. Animals are, in turn, associated with explicitly evaluative concepts such as [primitive] and [unsophisticated] (Saminaden, Loughnan, and Haslam 2010). Animals, particularly those raised for food, are also thought to be irrational-indeed, to lack a mind altogether (Bastian et al. 2012).
The lack of a mind has explicitly moral consequences. Bastian, Loughnan, Haslam, and Radke (2012) show that whether one has a mind strongly bears on whether or not we think of one as meriting moral consideration. Lynne Tirrell also notes the connection between animalization and diminished moral status. She convincingly argues that the Hutu's application of animalizing terms like "cockroach" and "snake" to Tutsi played a role in the 1993 Rwandan genocide (Tirrell 2012). Referring to Tutsi by these terms encouraged Hutu to see Tutsi as beings that are appropriately treated in the way one treats a cockroach or snake. Thinking of someone as an animal, particularly as a farmed animal, is also connected to thinking of that entity as having merely instrumental value, primarily for producing comestibles (e.g., milk, eggs) or constituting food (Bastian et al. 2012). Conceiving of something as having exclusively instrumental value, rather than as being an end in itself, clearly has consequences for moral consideration.
We have identified some biologizing conceptual connections that are morally problematic: the connections between having a biological or animal-like nature and lacking distinctively human characteristics like rationality or a mind at all, having merely instrumental value, and having a moral status. I turn now to the features of a perspective of a bodily part that are emblematic of fragmentation. In contrast with biologizing, thinking of someone as fragmented is to deny them characteristically, but not distinctively, human traits (Haslam 2006 (Bartky 1982;Nussbaum 1995;Langton 2009) identify as marks of the same phenomenon. 18 I have discussed biologizing and fragmenting features separately from each other, but there are important aspects of a perspective that reduce one to a part that we can only appreciate when we consider these aspects together-that is, when we consider the perspective that characterizes someone not only as a part but also as a bodily part. Components of a perspective on a bodily part are likely to be drawn from contexts in which a bodily part is highlighted, if not detached from the body. These include anatomy books, accounts of surgeries, and even experiences of seeing cuts of meat at a butcher shop. Indeed, meat-eaters, globally the majority of people, have substantial exposure to settings featuring nonhuman animals' bodily parts. For this reason, the perspective-taker is likely to think of the bodily part as meat. The concept of [meat] like that of [animal] is plausibly connected to [instrumentally valuable]. We also lack moral obligations towards meat. One might treat meat with impunity. This matches the way that people use "meat" to mark a lack of appropriate moral consideration, too. 19 With a clearer idea of the contents of a perspective on a bodily part and the connections among them, we can consider the process the hearer undergoes whereby she restructures her perspective on the female SUTW subject. As a perspective on the woman, the concept of the female subject is at its center. Which 18 The phenomenon is objectification. I remain agnostic here on the relationship between reducing to an anatomical part (biologizing and fragmenting) and objectification. 19 Another Harry Potter reference, this one uttered by Professor McGonagall in Harry Potter and the Goblet of Fire: "Potter is a boy, not a piece of meat!" Consider also the colloquial protest, "What am I, chopped liver?" properties would the restructured perspective on the woman have on my theory? Another way to consider this question is, which matches will a perspective-taker, the hearer of the SUTW, make between her perspective on the bodily part and on the woman?
The perspective on the woman the SUTW hearer takes is like the perspective on a bodily part: biologizing and fragmenting. The speaker reduces the woman to her body, and moreover, to a part of her body. Let us first see how it is biologizing: the woman's biological nature is "emphasized" or "highlighted." The perspective-taker connects the concept of the woman with ones like [purely biological], [exclusively bodily], and [animal]. These biologizing concepts figure centrally. Ortner (1974) notes that there is a pre-theoretic association between women and body, and between men and mind. There are a number of studies supporting the claim that people-both men and women-associate women with animals (Bernard et al. 2012;Heflick and Goldenberg 2009;Heflick et al. 2011;Loughnan, Haslam, and Bastian 2010;Reynolds and Haslam 2011;Rudman and Mescher 2012;Vaes, Paladino, and Puvia 2011). Irene López Rodríguez (2009) and Tipler and Ruscher (2019) argue that people connect women with animals, including those raised or hunted for their meat.
The perspective-taker also connects biologizing concepts with [instrumentally valuable], particularly [valuable for sex]. The indirect link between concepts like [exclusively biological] and [instrumentally valuable] is important. Recall that centrality is an explanatory relation: it is the woman's merely animal-like nature that explains, or justifies, exploiting her. In fact, the particular ways in which farmed female animals are exploited bear some affinities to one of the ways women typically are-namely, for their reproductive capacities. Farmed animals are manipulated to produce as many offspring and other by-products of reproduction (milk, eggs) as cheaply and efficiently as possible (C. Adams 2015). The widespread connection between consumption of flesh and sexual intercourse as "consumption" also suggests another parallel between the way animals and women are exploited (Hines 1999a(Hines , 1999b. 20 Farmed animals are exploited for people to "consume" their flesh, and women are exploited for men to "consume" sexually. Animals, especially wild animals hunted for their flesh, are also linked with romantic conquest and heterosexual sex (Kalof, Fitzgerald, and Baralt 2004).
I suggest that those fragmenting aspects of the perspective on the bodily part also shape the perspective-taker's view on the woman. The perspective-taker will connect her concept of the woman with concepts like [lacking subjectivity], [fungible], [interchangeable]. The woman's status as a part, like her biological nature, is central in the perspective: these components explain or justify her lack of subjectivity or disposability.
The biologizing and fragmenting aspects of the perspective, taken together, frame the woman as meat. Cognitive linguists Caitlin Hines (1999a) and Irene López Rodríguez (2009) attest to the association between women and meat. Some evidence is linguistic: the existence of certain classes of terms. One family of terms encompasses meat terms used synecdochically to refer to a woman as using an anatomical part characteristic of women. "Roastie," as in roast beef, is most frequently used among Involuntary Celibates. Another family encompasses terms that are often used to refer to characteristically female anatomical parts: "beef curtains" and "beef flaps" are examples. Based on an analysis of language used to refer to women, people also connect women with a related conceptual domain: food (Hines 1999a(Hines , 1999b.
I have focused on the cognitive aspects of the perspective on the female subject restructured by comparison with a perspective on the bodily part, but the perspective also has affective dimensions. In taking on a perspective with those elements associated with meat, especially if those components are centrally embedded, the perspective presents the woman in a way that suggests that certain attitudes and emotions that are appropriate for meat are appropriate to have towards the woman. Ethically salient attitudes include callousness, indifference, or an "instrumental" attitude-that is, seeing the woman as something to be used however, and discarded whenever, one likes.
The hearer's perspective need not contain all of or even any particular one of these components, cognitive or affective, but it should comprise a sufficient number such that she thinks of the woman in a way that emphasizes her body, as lacking traits traditionally associated with humanity (e.g., rationality), and as having a diminished moral status in virtue of these traits. We should also note that characterizing a woman as, say, lacking subjectivity or as irrational does not entail that the hearer believes this about the woman. The perspective-taker has only engaged in a kind of thought that frames the woman in a way (namely, as a bodily part) that would make the imputation of these traits apt. A hearer need not believe that the woman is disposable any more than the hearer of "Pettigrew is a rat" need believe that Pettigrew is sneaky. This is not to suggest that perspectives lack the power to shape the perspective-taker's beliefs. Indeed, there is an especial danger that this might occur. As the research in social and cognitive psychology and cognitive linguistics that I cite attests, people are already disposed to associate women with entities (animals, meat) that undermine their humanity.
We should pause to emphasize the nature of perspectives. They are neither wholly cognitive nor entirely noncognitive. They contain concepts but are affective in various ways. Some of the constituent concepts, like those capturing thick terms, are partly affective. Connecting certain elements of a perspective also inspires one to have certain attitudes or emotional responses towards the person on whom the perspective is taken. Perspectives, according to Camp (2017), are not reducible to propositions. Yet, one can articulate one's perspective in propositional terms. A hearer of (1) could begin to articulate the perspective she takes on Ivanka in terms of propositions: that she is only valuable for her appearance, that she is not intelligent, that she lacks a mind of her own, and so on. I say "begin" for one cannot fully or exhaustively articulate a perspective in propositional terms. Perspectives are ultimately ineffable. 21 I have argued that taking a perspective on a woman effectively models the type of thought that one experiences in hearing an SUTW. My next task, which I undertake in section 6, is to explain how my view accounts for the data ( §3.1-3.4). In the next section, I also explain how SUTWs are distinctive from non-SUTW utterances like (4), which might seem amenable to the same analysis.

How a Perspectival Theory Explains the Data
We should be clear as to how my theory explains the data in favor of thinking that an SUTW speaker undertakes a derogatory act or does something other than convey the sentence it literally expresses ( §3). The SUTW speaker does not merely assert the literal significance of the sentence uttered but also enables the hearer to take a perspective on the female subject of that sentence that reduces her to the bodily part. Taking on such a perspective also explains the hearer's experience of offense and the felt sense that the SUTW target has been reduced. A hearer is offended because the perspective that she takes on the female subject presents her in a biologizing and fragmenting way. I can also explain why the SUTW speaker predicates an anatomical part term of the female subject as opposed to other language. Terms like "piece of ass," by virtue of being bodily, and bodily parts, plausibly evoke different associations than superficially similar language.
The perspectival theory presented here also explains data pertaining to the type of act an SUTW constitutes ( §4). Understanding the pattern of associational thinking in which the hearer engages as taking a perspective on the female subject predicts the cognitive-seeming behavior of the SUTW content. A perspective contains cognitive elements, and it can be at least partly articulated propositionally. This explains why some explications are apt, how a party to the conversation can refer to what the speaker says, and how someone can (dis)agree with the speaker. I can explain, as well, the data that weighed against a cognitive analysis of the content the SUTW speaker conveys. Although one can articulate apt paraphrases, one cannot exhaustively capture all that is contained within the perspective in propositional terms. A perspective is ultimately ineffable. One reason is because some of the elements composing it are, plausibly, propositionally inarticulable: evaluatively imbued properties or the feelings the hearer experiences in taking on the perspective. Another reason is that no propositional articulation could express the "fitting-ness" relationship among the perspectival elements. Pace Camp (2017), I am also skeptical as to whether a propositional articulation of a perspective could capture the centrality and prominence relations among that perspective's contents.
A speaker is able to coherently deny any particular paraphrase because the perspective that the hearer takes on, which the objector propositionally (partially) articulates, need not have any one particular content. No one of the components distinctive to a biologizing and fragmenting perspective must be a part of the perspective the hearer takes on.
Two other features to be accounted for include two kinds of contextual sensitivity: the aptness of some paraphrase and a speaker's ability to plausibly deny some paraphrase. These are connected, for the extent to which a paraphrase is apt to a hearer is directly related to how plausible the speaker's denial of conveying it will be to that hearer. For this reason, the feature of a perspective that explains the contextual variability of aptness should also explain the ability of a speaker to more or less plausibly deny some paraphrase, depending on the context.
It is easier to see how perspectives handle context-dependence by considering an example. Suppose that the hearer of (1) takes a perspective on Ivanka containing [valuable for her body]. The hearer can paraphrase what the speaker conveys as "that Ivanka is valuable for her body." The degree to which the paraphrase is apt in a given context corresponds to how central or prominent the perspectival content is. 22 Thus, if [valuable for her body] figures centrally or prominently in the hearer's perspective on Ivanka, "that Ivanka is valuable for her body" will seem apt. If [disposable] figures less centrally or prominently in the hearer's perspective, "that Ivanka is disposable" will strike the hearer as less apt. Now, a perspectival component's position within a perspective varies with the context of utterance. The same utterance in different mouths, or uttered on different occasions, or aimed at a different target can result in a different structuring of perspectival contents. That is, some contents might figure more or less centrally or prominently based on the context. For instance, what makes (1.1P) more apt as a paraphrase of (1) in a context like that which Lovelace describes is that the contents on which that paraphrase is based figure more centrally or prominently in that context of utterance as opposed to other situations.
At this point, a question emerges. It might seem that non-SUTWs like (4) are acts that can be explained in the same way that I explain SUTWs: as acts (a) whereby the speaker conveys a perspective and (b) a perspective that characterizes the female subject as meat. One could see why someone might be tempted to explain (4) in the same way: (4) involves a female subject and a bodily part characteristic of female bodies. It is also (in many contexts) derogatory and causes offense despite the fact that the proposition (4) literally expresses is not derogatory. This leads one to think that the speaker performs an act other than (if not, in addition to) asserting the proposition literally expressed, which is also true of SUTWs. We can call this act the "nonliteral" act. If it is true that (4) could be given the same analysis as SUTWs, SUTWs would not be distinctive from utterances like (4), a claim that also has consequences for the especial social import I claim SUTWs have in section 7. 23 Utterances like (4), however, are not plausibly analyzed in the way SUTWs are. It is doubtful that (4) evokes a perspective at all, let alone one in which a woman is both biologized and fragmented. Firstly, a hearer of (4) does not undergo perspectivetaking as described in section 5. When we interpret (4), we do not experience anything like seeing the woman as a pair of legs as we do when someone identifies the female subject with "pair of legs." We do not experience a restructuring of our perspective of the woman by comparing its contents with the contents of our perspective on women's legs. Aside from our interpretive experience, thinking of the woman in terms of the bodily part is the wrong kind of interpretive process. The speaker of (4) literally says that the woman "got" legs-not that she is identical with her legs. An interpretive process whereby the speaker tries to, essentially, discern which properties a woman and her legs share mistakes a claim about possession with a claim about identity.
Even if we grant that the speaker of (4) prompts the hearer to take on a perspective, it is unlikely that it is the same kind of perspective as that conveyed by an SUTW. The perspective that a hearer of (4) takes on might be biologizing since the speaker attributes the possession of a bodily part associated with women to the female subject. Yet, if a perspective were associated with (4), it is not plausibly fragmenting-and, so, cannot be one in which the woman is characterized as a piece of meat. Objecting to the speaker of (4) using (4.1) or (4.2), for instance, is infelicitous, and the articulations of (4) these objections and (1.3) contain are inapt.
(4.1) Are you saying that she's just a pair of legs? (4.2) Are you saying that all that's important about her is her legs?
Yet, one could use (4.1) or (4.2)-or the content of (1.3)-to object to (2) without oddness and claim that (4.1), (4.2), and (1.3) are plausible articulations of what the SUTW conveyed. Furthermore, in hearing (4), we might be offended for a number of reasons, one of which might be that saying that someone possesses a certain bodily part draws attention to. Yet, we do not feel that the woman has been reduced to the part. Our reactions, like the claims that one use to paraphrase or object to utterances like (4), are sensitive to the distinction between a claim about possession and one about identity.

The Social Importance of Synecdochical Utterances Targeting Women
I have offered a theory of what an SUTW speaker does: she prompts the hearer to think of the female subject as reduced to a bodily part, which is to think of her in a way that is biologizing and fragmenting-in short, as a piece of meat. I argued that this kind of thinking is a perspective. A perspective fits both those features, suggesting that the content that is cognitive and noncognitive. In closing, we should reflect on the possible threat SUTWs pose to an egalitarian society and, in so doing, see how SUTWs fit with extant work on other kinds of socially salient speech.
The content that an SUTW speaker invites the hearer to engage with-a perspective on a woman as a mere piece of meat-is, clearly, extraordinarily dehumanizing. In a society in which the presentation of women as significant only in virtue of their bodies is prevalent, it is plausible that SUTWs run the risk of reinforcing, perpetuating, and legitimating these representations. This concern is even more pressing given a feature Camp identifies as characteristic of perspectives: irresistibility. In hearing an SUTW, one automatically takes on the perspective (Camp 2017). The hearer need not consciously entertain the way in which the woman and the bodily part are alike. To think as the SUTW speaker does is, in turn, to be cognitively complicit in her reprehensible thinking (Camp 2013).
To accurately gauge how pernicious SUTWs are, we need a theory of how the biologizing and fragmenting perspective interacts with larger social structures either directly or via mechanisms like David Lewis's (1979) conversational scoreboard or Robert Stalnaker's (2002) common ground. If SUTWs are oppressive exercitives (McGowan 2009), then using them constitutes the use (and with every use, a strengthening) of an abstract oppressive norm; and most chillingly, this occurs irrespective of whether the interlocutors are aware that their speech does so. Langton (2012) has also proposed amendments to models of the common ground and the conversational scoreboard amenable to nonpropositional content. Her model might hold promise for explaining how perspectives might be presupposed among conversational participants. In that project, Langton seeks to understand how hate speech and pornography can encourage (in the sense of causing and advocating) hearers to alter their desires and attitudes. Langton's theory might also explain how an SUTW might perpetuate biologizing and fragmenting views of women by altering hearers' conative dispositions.
One of the most disturbing features of an SUTW's derogatory content is its "slipperiness." Some of its content cannot be articulated propositionally. Even when it can, it is indeterminate and contextually variant-and a speaker can deny conveying any one paraphrase. These "slippery" features create the illusion that no derogatory content was conveyed at all, producing a "gaslighting" effect on some hearers. For these reasons, the most important work on SUTWs will consider strategies for calling out speakers on what they convey in uttering SUTWs and for resisting and opposing the perspectives forced upon us.
AMANDA McMULLEN is an assistant professor of philosophy at the University of Arkansas, Fayetteville. She specializes in philosophy of language and feminist philosophy. Currently, her work focuses on types of socially significant speech that function to disparage, or otherwise harm, marginalized groups.