What a jerk!

I argue that “ general pejoratives ” such as “ jerk ” or “ bastard ” differ crucially from items such as “ that damn N ” . While items such as the latter typically serve to give vent to one's attitudes, general pejoratives essentially involve judgments about a person's behaviour or character. This is particularly evident in cases where pejoratives occur not as epithets, but as predicate nominals. If we want to account for the overall contribution of words such as “ jerk ” , there are three kinds of content that ought to be distinguished: truth-conditional contents, evaluative presuppositions, and expressive contents that are either at-issue (in the case of expressive predicates) or non-at-issue (in the case of epithets).


| INTRODUCTION
Over the last twenty years or so, scholars have become increasingly interested in expressive speech.Here are two well-known examples: (1-1) This cur howled the whole night.(Frege, 1897: 152) (1-2) That bastard Kaplan was promoted.(Kaplan, 1999, p. 12) 1 A particularly salient feature of expressive items is their projection behaviour: whatever attitude is conveyed by (1-1) or (1-2) is also conveyed by the negation of these sentences.This observation has led to the common view that expressives are examples of a broader family of "non-at-issue" or "projective" contents.Potts (2005) and Horn (2007Horn ( , 2013) ) see them as examples of conventional implicature (but mean different things by that term); Schlenker (2007) construes them as presupposition triggers; and McCready (2010) as well as Gutzmann (2021) employ a category dubbed "CIE-content", thereby leaving it open whether conventional implicatures and expressives are examples of the very same phenomenon.Now compare (1-1) and (1-2) to the following sentences: (1-3) Fido is a cur.
Intuitively, what appears in (1-2) as a secondary content reappears in (1-4) as the speaker's central message (and vice versa), and assuming that the speaker of (1-1) is talking about Fido when referring to "this dog", (1-3) may be taken to express exactly the same emotional attitude as (1-1).But there is a crucial difference between these two pairs.When negating either (1-3) or (1-4), the expressive contribution of "cur" and "bastard" does not project; instead it simply seems to vanish.When denying that A is a cur or that B is a bastard, one does not seem to express anything (or, in any case, not an attitude towards A or B).
This paper is concerned not with expressives in general, but with a specific sub-type that is exemplified by "bastard". 2Such words have been discussed under a variety of monikers: "particularistic pejoratives" (Saka, 2007, p. 148), "general pejorative terms" (Hay 2011), and "personal slurs" (Bach 2018).Throughout this paper, I will simply call them "pejoratives".What makes these words particularly intriguing is that they can be used as parts of expressive epithets ("That bastard NN is F") but also as predicate nominals ("NN is a bastard"), which naturally leads to the question of how these uses are related. 3(In what follows, I will sometimes use the term "p-pejorative" for the latter case, and "e-pejorative" for the former.) According to some theories, there cannot be a strictly uniform account of pejoratives.Some authors have claimed, for instance, that the epithet "that bastard" is a genuinely expressive item whereas the predicate "… is a bastard" contributes ordinary at-issue content.One reason for not taking that route is offered by Geurts (2007, p. 210), who points to sentences such as: (1-5) That bastard Schmidt is a bastard.
(1-6) That bastard Schmidt is not a bastard. 4  The first one looks like a pleonasm, and the second like a blatant contradiction.Assuming that the two occurrences of "bastard" in each sentence function in vastly different ways will make it very hard to explain this observation.
The main aim of this paper is to offer a theory of pejoratives that can account for both types of use, and here is how I will proceed.I will start, in Section 2, with a brief overview of the theories currently on the market.Section 3 discusses some descriptive constraints on using pejoratives.In Section 4, I turn to the fundamental question of whether pejoratives have neutral counterparts, and in Section 5, I present my own account according to which pejoratives carry three kinds of content at once: truth-conditional contents, evaluative presuppositions, and expressive contents that are either at-issue (in the case of expressive predicates) or non-at-issue (in the case of epithets).
seems, on the one hand, like a paradigmatic expression of an emotion.How could one call someone a bastard without thereby expressing one's attitude towards that person?On the other hand, the expressive contribution of "bastard", as it is used here, appears to be sensitive to the effects of entailment-cancelling operators such as negation.
Seen from this angle, the contribution of "bastard" may appear purely descriptive.How can we account for these two features that push us in different (and seemingly incompatible) directions?
Let us start with a brief overview of the few extant accounts.We shall see that many of them simply ignore one of these two features, thereby assimilating p-pejoratives either to ordinary at-issue contents or to purely expressive contents.
The former option is embraced by Rett and Potts.On Rett's view, the word "bastard", as it is used in (2-1), "appears to contribute descriptive content, as negating or denying the sentence does in fact result in the negation or denial of [Chapman's] status as a bastard" (Rett, 2021, p. 194).Similarly, Potts (2007, p. 194) has it that "predicates that appear in copular position must necessarily fail to be expressive".Let us call such a view "pure descriptivism".
While Rett and Potts claim that a sentence such as (2-1) altogether lacks expressive content, Gutzmann has advanced (roughly) the opposite claim.On his view, in sentences such as (2-1), "the expression of a negative attitude towards [Chapman] seems to be the main point" (Gutzmann, 2015, p. 270;cf. Kaplan 1999, p. 27).Accordingly, (2-1) would be an example of an expressive (or "use-conditional") content that is at-issue and that lacks truth-conditional content proper.Let us call this view "pure expressivism". 5  Finally, there are scholars such as Richard (2008, p. 34), Hay (2011), Lasersohn (2017) and Berškytė (2021), who have advanced hybrid accounts.According to such accounts, pejoratives involve a mixture between truth-conditional and expressive content.For instance, on Lasersohn's (2017, p. 233) view, the truth-conditions of "John is an asshole" are "roughly the same as those of John is obnoxious, though the former sentence packs a stronger emotional punch".
One of the essential aims of this paper is to show that hybrid accounts are on the right track, but not yet complex enough.Instead of a two-dimensional semantics we need a three-dimensional one.But let us first turn to the question of whether the use of pejoratives is subject to descriptive constraints at all.If the answer to this question is yes, then pure expressivism can be safely discarded.

| DESCRIPTIVE CONSTRAINTS ON USING PEJORATIVES
In a recent empirical study on pejoratives, Cepollaro et al. presented Italian speakers with several versions of a brief story involving the Italian pejorative "stronza" (or its male counterpart "stronzo").The versions differed, first, in either containing or not containing an additional sentence (appearing below in angle brackets), and second by employing either an epithet or an expressive predicate (the two options in the curly brackets below).Here is the English translation, where "stronza" has been rendered as "jerk": Pietro and Marta are waiters in a restaurant.Bea is the assistant chef.<Her colleagues do not like her much because she is rude to everyone.>When Bea finishes her shift, Marta tells Pietro: {Bea is a jerk.
(Bea è stronza)/That jerk Bea (Quella stronza di Bea …) has finished.}(Cepollaro et al., 2021, p. 9282, presentation slightly modified) The participants were then asked to rate the "acceptability" of Marta's utterance on a scale ranging from 1 (unacceptable) to 5 (completely acceptable).Importantly, the notion of acceptability was not explained any further to the participants.As a consequence, the participants may, as Cepollaro et al. (ibid.,9283)themselves concede, have rated not the pragmatic felicity, but instead things like social acceptability, but I will not pursue this issue any further here.
The main result was that in the case of both e-pejoratives and p-pejoratives, the additional sentence (in angle brackets) increased the acceptability rating by, roughly, 1.5 points, resulting in a rating of around 4. What is the lesson to be drawn from that experiment?According to Cepollaro et al., the experiment suggests that a wide-spread view of pejoratives is flawed.This wide-spread view can be best characterised by comparing expressives to some other items that carry non-at-issue contents: (3-1) That bastard Chapman got promoted.
(3-2) Sam is having dinner in New York tonight, too.(Kripke, 2009, p. 373) (3-3) My aunt's cousin went to that concert.(Grice, 1989, p. 274) (3-2) and (3-3) contain items that are standardly classified as presupposition triggers: by uttering (3-2) one presupposes that some other person is having dinner in New York; by uttering (3-3) speakers presuppose that they have an aunt (etc.).But there is a crucial difference between these two examples.(3-3) is an example of an accommodable presupposition. 6If the speaker's interlocutors are not aware that the speaker has an aunt (etc.), the utterance is still felicitous.By contrast, the mechanism of accommodation does not seem to be available in the case of (3-2).The information that someone else is having dinner in New York must be contextually salient when that sentence is uttered.Following Tonhauser et al. (2013), I shall call this specific constraint on items such as "too" strong contextual felicity (SCF).
Most scholars agree that expressives are not subject to the SCF constraint (cf.Potts, 2005, p. 33;Schlenker, 2007, p. 240;Tonhauser et al., 2013, p. 76).You may felicitously utter (3-1) even when your addressees are ignorant about your attitude towards Chapman.Now according to Cepollaro et al., the above experiment casts doubt on this view.They acknowledge (2021, p. 9280), however, that the experiment does not directly test whether there is an SCF constraint on pejoratives.Since the different versions of the story differ essentially in whether they include the sentence "Her colleagues do not like her much because she is rude to everyone", what is tested directly is rather whether additional information as to i. the behaviour of the target and ii. the speaker's (or, more generally, all of the interlocutors') attitude towards the target increase the acceptability of sentences involving "stronza". 7 If the mainstream view is characterised as the claim that items such as "stronza" do not "impose any contextual constraints" (Cepollaro et al., 2021, p. 6279, emphasis added), then the experiment indeed casts doubt on that view.
However, most "mainstream" authors seem to be concerned with a more specific constraint, namely with whether an utterance involving pejoratives is felicitous only if the interlocutors are already aware of the attitude held by the speaker.That there is no such constraint has been experimentally confirmed in the case of Guaraní pejoratives (Tonhauser et al., 2013, p. 76), and the same seems to be true of English: when speakers employ words such as "jerk", they routinely express their attitudes in a potentially "informative" way.
So the experimental study conducted by Cepollaro et al. does not refute the mainstream view, but I think it successfully demonstrates that there are some contextual constraints that have often been unjustly ignored.If an additional sentence that characterizes a person as "rude" increases the acceptability of sentences involving the pejorative "stronza", then one might say that the rudeness of a person (or some other negative character trait) gives us a positive reason for calling someone a jerk, which may be taken to show that pejoratives cannot be given a purely expressivist analysis.And one might even claim that words such as "rude" specify the descriptive content of pejoratives (more on which in the next section).Now importantly, the idea that pejoratives are part of the "game of giving and asking for reasons" is plausible even if there were a flaw in the above experiment.Speakers regularly offer reasons for calling somebody a jerk, and typically they do so in a narrative manner, focussing on the behaviour of a particular person.Consider an example: I turned to see a tall bald man looking down at me as the train pulled in to the platform.I let two people in before me, and that's when I felt the push.As we turned toward the seats I felt another push on my back, and again looked at the man, who now released an annoyed huff of breath.What a jerk!I thought.Does he think that he's the only one who deserves a seat? 8 The author's point seems to be that the tall bald man is a jerk because of behaving in a particular way.In this respect, pejoratives are crucially different from purely expressive items such as "that damn".In "That damn Chapman got promoted because he did so-and-so", Chapman's having done so-and-so can only be understood as a reason for his promotion, not as a reason for holding the expressed attitude.The utterance "Chapman is a bastard because he did so-and-so", by contrast, seems to rationalize the ascription of a partly descriptive property.

| DO PEJORATIVES HAVE NEUTRAL COUNTERPARTS?
In the introduction, I offered two examples of expressives, "bastard" and "cur".Now there is an apparent difference between these items: the expressive force of the latter item can easily be detached (yielding the neutral term "dog"), while in the former case it is difficult to find a neutral counterpart (NC) that specifies the term's descriptive content.
This may lead to the idea that there is no such thing as an NC, for one reason or other.One possible reason would be that pejoratives altogether lack descriptive content.This claim has been advanced by Kaplan who writes that "the epithet 'that bastard' has only expressive, and no descriptive force" (1999, p. 27).If this view is on the right track, then some form of pure expressivism will be the most natural theoretical option.However, an account of that kind would make it rather difficult to accommodate predicative uses.Now one thing is clear: there is no obvious NC to the word "bastard".But the inference from "There is no straightforward NC" to "There is no descriptive content whatsoever" seems invalid.Furthermore, there are two features of pejoratives that offer a (partial) explanation for why there is no descriptive term that quickly springs to mind as a genuinely descriptive counterpart to "bastard".The first feature is well-known from the debate on contextualism.Understanding what is specifically meant by a word or a sentence will often require a good deal of knowledge about the context of utterance, and pejoratives are plausibly subject to essentially the same kind of context dependence.The descriptive standards for being a jerk are, as Hay (2011, p. 459) has suggested, different depending on whether the term is used in a sentence such as "That driver is a jerk" or when used in a romantic context ("Don't date him.He's a jerk").So in this respect, the difficulty of specifying what is meant by "jerk" may not be crucially different from the difficulties surrounding sentences such as "Tipper is ready" (Bach, 1994, p. 130).If no context is specified, sentences involving "ready" may invite the response "For what?", and sentences involving "jerk" may lead to the question "In what respect exactly?" The second feature is more specific to pejoratives.Though ordinary speakers may not be able to spontaneously offer a descriptive paraphrase of words such as "jerk", they may on reflection come up with adjectives such as "disagreeable" or "obnoxious".And here one might argue that such words aren't NCsfor the simple reason that they are, like "jerk" itself, loaded words.More specifically, they may be construed as "thick terms", i.e. as terms that "somehow 'hold together' evaluation and nonevaluative description" (Väyrynen, 2013, p. 2).Accordingly, one might claim that there is, in general, no genuinely descriptive paraphrase for words like "jerk".
I think this is true, but there seems to be a rather obvious explanation for this observation, namely the mechanism that has come to be known as the "euphemism treadmill".Suppose you live in a society in which, say, talking about human reproductive organs is deemed inappropriate.In such a society, any euphemism intended as a replacement for an extant vulgar term will tend over time to become as vulgar as the original one.The vulgarity is, after all, grounded in (the perception of) a certain subject matter.Similarly, when there are, within a community, stable standards for inappropriate behaviour, then any term intended to merely describe such behaviour may tend over time to acquire a negative overtone.
WHAT A JERK!So the familiar phenomenon of context-dependence offers a partial explanation for why it is difficult to pin down the exact descriptive contribution of "jerk", and the euphemism treadmill explains why possible NCs are often not strictly neutral.Now assuming that pejoratives have some kind of descriptive content, then what exactly might it be?Here are the two OED entries for "jerk" and "bastard", both of which draw heavily on thick terms: (jerk) Now: an objectionable or obnoxious person.Usually with reference to a male.
(bastard) Used as a term of abuse or contempt for a person (esp.a man or boy), now esp.for someone who is callous or wilfully cruel, or who acts ruthlessly out of self-interest. 9  The OED makes a somewhat peculiar distinction here: while "jerk" is directly defined via a disjunction, "bastard" is presented as a "term of abuse" that is more loosely linked to character traits such as ruthlessness.But I do not think that we ought to take that distinction too seriously, especially since the Concise OED (12th edition) simply defines "bastard" as "an unpleasant or despicable person".And neither should we worry too much about the specific adjectives.Even mundane words like "sandwich" are difficult to analyse, so we clearly shouldn't expect "jerk" to be straightforwardly definable via a short disjunction.But perhaps a longer disjunction might do the trick.If we use "NC" as an index which indicates that any non-descriptive aspect of meaning, be it expressive or evaluative, has been detached, then one might attempt to specify the descriptive content of the two above pejoratives as follows: (Def.jerk) x is a jerk NC iff x is objectionable NC or obnoxious NC or … (Def.bastard) x is a bastard NC iff x is unpleasant NC or despicable NC or … And there is one observation that speaks in favour of such an analysis.Consider the following sentence, uttered by Tom: (4-1) Jerry is a jerk, but he is neither unpleasant nor despicable nor … Of course, someone may be a jerk without being either unpleasant or despicable (full stop), but suppose Tom adds more and more adjectives.Then we might get the impression that he does not really know what he is saying.One cannot call someone a jerk without assuming that that person has some negative property. 10  Now the test I have just employed is a version of Grice's (1989, p. 44) cancellability test, and while that test may be a decently reliable method for identifying conversational implicatures, it clearly cannot distinguish between semantic content and things such as felicity conditions. 11So one might concede that adjectives such as "obnoxious" are somehow related to the felicitous use of "jerk", but at the same time maintain that do not they not specify the descriptive content of that word.Kaplan's account provides an illustration of such a view.He concedes that "bastard" is not an all-purpose pejorative (for example, it cannot be used as a "derogation for being weak"), while denying that the epithet "that bastard N" has any descriptive content.He continues thus: The mere fact that there are usage rules for a particular epithet that confine its correct use to a specialized range of cases does not show that it has descriptive content […].I may choose a specialized instrument to express my attitude, based on my belief that the target has certain characteristics, without asserting that the target has those characteristics […].This is easily seen in the honorifics "Sir" and "Ma'am", forms of address for, respectively, males and females, that express respect, but not respect for being, respectively, male or female.(Kaplan, 1999, p. 27) Kaplan's idea can be illustrated by an example where there are very clear-cut descriptive standards.The felicitous use of the honorific "Your Eminence" is restricted to cardinals of the Catholic Church, which makes it rather exclusive: at present (in July 2024), only 236 people can be addressed in that way.But intuitively, an utterance of (4-2) I am bitterly disappointed, your Eminence.proper?In the literature on non-at-issue contents, there is the common assumption that the secondary status of some content can be tested by the "Wait a minute" diagnostic (cf.Shanon, 1976, pp. 248-9, Lewis, 1979, p. 339).In the case of "Your Eminence", the test works as intended: an utterance such as (4-2) can be challenged by saying "Wait a minute, am I really a cardinal?",but not by saying "That's false, I am not a cardinal".Now unfortunately, this diagnostic is insensitive to the very difference Kaplan is concerned with.Just as the cancellability test does not tell us whether some piece of information is a felicity constraint, an entailment or a conventional implicature, so exclamations such as "wait a minute" can be used to "object to a wide range of non-at-issue content, including conventional implicatures, appropriateness conditions, and conversational implicatures" (Potts, 2015, p. 174-5).The test is ultimately a test for whether something is not the main point of an utterance.
But fortunately, the "reverse" diagnostic is helpful here.On the assumption that standard denials target only atissue contents (cf.Potts, 2015, p. 174), the felicity of the following dialogue (4-3) A: Chapman is a bastard.
B: That's not true; he is not a bastard.
would show that p-pejoratives contribute at-issue content.And if we further assume that "bastard" always has essentially the same meaning, regardless of whether it occurs as a predicate noun or as part of an epithet, then we can argue as follows against Kaplan's claim: (P1) The sentence "A is not a bastard" is (or involves) an ordinary denial of a descriptive claim.
(P2) The word "bastard", uttered as part of an epithet, does not mean anything different than is its wont when it occurs as a predicate nominal. 13  (C) Therefore, the descriptive constraints on the epithet "that bastard" are a matter of descriptive content too.
Let us now turn to the opposite problem.Does "bastard" contribute expressive content?With respect to epithets, the answer is unanimously yes, but quite a few scholars think that p-pejoratives are examples of ordinary at-issue content.Such a view may come in two flavours, which I shall call, respectively, the unabashed and the conciliatory version.According to the unabashed version, a typical utterance of (4-5) Chapman is a jerk.
simply makes a claim about one of Chapman's properties and is not used to express a non-doxastic attitude at all.
Such a claim invites an incredulous stare as a reply.That speakers are typically conveying an emotional attitude towards a particular person when uttering sentences like (4-5) seems like a truism.
This brings us to the conciliatory version.Potts's (2007, p. 194) claim that "predicates that appear in copular position must necessarily fail to be expressive" can be made compatible with the above intuition by assuming that the emotive extra content is carried by something like a conversational implicature.There are, however, two problems with this strategy.First, it is not easy to see how the Gricean mechanism might account for such (alleged) implicatures.The very notion of implicature is typically restricted to cases where the implicatum is a propositional content.So does the idea of there being expressive implicata even make sense? 14Second, even if a Gricean explanation were available, it would commit us to the view that the expressive content conveyed by (4-5) is cancellable.But this does not seem to be the case, as witnessed by the following example: (4-5a) Chapman is a jerk, but I don't hold a negative attitude towards him.Echoing a remark of Williamson's (2009, p. 150), we might say that an utterance of (4-5a) merely adds hypocrisy to hostility.
So if what I have argued thus far is on the right track, then much speaks in favour of a hybrid account according to which the contribution of pejoratives involves both descriptive and expressive contents.(Notice that this assumption does not yet rule out that one of these contents is, in some sense or other, not at issue.) Let me end this section with a list of the desiderata for a successful account of pejoratives.All of these desiderata have already been mentioned before, but in what follows I have chosen the wording so as to remain as noncommittal as possible.
(O1) A freestanding occurrence of "A is a jerk" carries an emotive punch.
(O2) By uttering "A is not a jerk", one does not express an emotive attitude towards A (but perhaps an attitude of a more general kind).
(O3) The felicitous use of pejoratives is subject to certain descriptive standards (though these standards may be rather vague).
(O4) There are semantic links between e-pejoratives and p-pejoratives.Part of the content of "That bastard A is F" can be challenged by uttering "(Whadda ya mean), A is not a bastard".

| UNDERSTANDING EXPRESSIVE PREDICATE NOMINALS
The four observations from the previous section speak in favour of an analysis of "jerk" (and its relatives) that is more nuanced than the extant accounts.Most importantly, we should first reject the wide-spread assumption that such words are "particularistic" (Saka, 2007, p. 148), i.e. solely directed at an individual, not at an entire class of persons.And second, we ought to construe pejoratives as a means of making a kind of moral judgment and not just as a way of conveying one's attitude. 15  The moral evaluation inherent in words such as "jerk" has been recently stressed by some scholars who have applied their analytical skills to pejoratives.Consider two such proposals: [The] asshole is the guy … who systematically allows himself special advantages in social relationships out of an entrenched (and mistaken) sense of entitlement that immunizes him against the complaints of other people.(James, 2016, p. 41) [The] jerk culpably fails to appreciate the perspectives of others around him, treating them as tools to be manipulated or idiots to be dealt with rather than as moral and epistemic peers.

(Schwitzgebel, 2014, emphasis removed)
There is reason to be skeptical about some details of these two analyses, but the details are not important here.
What is important are two points.First, there are descriptive standards for whether a person is, say, a jerk, and these standards have to do with a person's character or behaviour. 16Second, the two words "asshole" and "jerk" are not descriptively equivalent; speakers make subtle distinctions between them.Consider the following fictional story, due to Nunberg: On Sept. 11, 2001, with all flights cancelled across the country, you're in the Hertz rental agency in Manhattan, trying desperately to rent a car to get home to your family in Texas, along with a large crowd of anxious people trying to do the same thing.A man walks in, pushes to the front of the crowd, and asks the clerk, "Where's the Hertz Gold Card line?"You turn to your friend standing next to you and say, "What a(n) ___!" (Nunberg 2012, p. xii).Nunberg asked a few dozen people which pejorative best fits the situation.Almost all of them opted for "asshole", but they generally preferred words such as "bastard" or "douchebag" when being asked about other scenarios.
There is yet another observation that speaks in favour of an essentially behaviour-based account of pejoratives.
Consider the following examples, the first of which I have already quoted in Section 3: (5-1) What a jerk!
(5-3) You are the biggest jerk I've ever met in my entire life. 17  What these examples have in common is that all of them seem to involve the idea that being a jerk is a matter of degree.This is quite evident in the case of , which explicitly says that the addressee is not just an ordinary jerk.
One might claim of course that a sentence such as (5-2) differs from an ordinary attribution of jerkness ("A is a jerk") only in that (5-2) carries a stronger emotive oomph, but on the plausible assumption that "such" has in  the same function that it has in, say, "It was such an inconvenience!" (Bolinger, 1972, p. 69), an utterer of (5-2) seems committed to there being varying degrees of being a jerk.
In the remainder of this section, I will offer an account of pejoratives that can fulfil the desiderata listed at the end of Section 4 and that is also well poised to accommodate the view that being a jerk is not a black or white issue.
(However, to keep things simple I will focus on examples that do not involve any degree words.)We shall see that in order to fully understand pejoratives we need to distinguish three layers of content.An analogy may help to see the point.In the cross examination of Oscar Wilde, Carson (the attorney defending the Marquess of Queensberry) asked Wilde several times whether he thought a certain story to be blasphemous.Wilde initially attempted to evade the question by saying things such as "It was badly written" or "I thought it disgusting", but ultimately resorted to the strategy of semantic ascent: CARSON."Do you consider that [quote from the story] blasphemous?"WILDE."I think it is horrible -'blasphemous' is not a word of mine."(Foldy, 1997, p. 8) Why did Wilde refuse to say "The story is (not) blasphemous"?The general outline of the answer seems obvious: "blasphemous", like other thick terms, is fraught with ought, and the "ought" projects even when one is denying that something is blasphemous.In using (as opposed to mentioning) that word, one almost inevitably seems to subscribe to certain moral standards 19 standards Wilde did not share.Note that this kind of projectivity is crucially different from the one scholars have focussed on when discussing expressives.The sentence "That bastard Chapman did not get promoted" expresses exactly the same attitude towards Chapman as its unnegated counterpart.In contrast, the sentence "The story is not blasphemous" does not convey an attitude towards the story, but the sentence is still committal with respect to particular "values".
Now the important thing is that much the same could be said about p-pejoratives.Compare a now well-worn example: (5-4) Chapman is not a bastard.
By uttering (5-4) one is clearly not expressing a negative attitude towards Chapman, but if the analogy between thick terms and pejoratives is valid, then  would convey an attitude of a more general kind.In order to count as a bastard, someone has to meet a particular descriptive requirement F, and even (5-4) seems to convey a negative Kyle: Conner is not a bastard.(Potts, 2005, p. 157) seems to presuppose the existence of certain standards of bastardness as well as the assumption that meeting these standards is generally a bad thing.The main point of Kyle's utterance seems to be that Conner simply does not meet these standards.20   So in order to understand the expressive or evaluative contribution of words like "bastard" (or "blasphemous"), we ought to keep three things apart: a.An attitude towards a particular person (or object) b.The criteria that have to be met in order for something to deserve that attitude c.A general attitude towards things of a certain kind Importantly, not all of these things must be part of the actual communicative content of utterances involving pejoratives.An analogy might be helpful again.In contemporary meta-ethics, quite a few scholars have claimed that in using "thin" moral concepts such as "good" we express attitudes towards things of a certain kind.For instance, on Ridge's view, the notion of moral reason ought to be analysed as follows: "There is moral reason to X" expresses (a) an attitude of approval of a certain kind toward actions insofar as they have a certain property and (b) a belief that X has that property.(Ridge, 2006, p. 315) On that view, what is expressed by the sentence (5)(6) There is moral reason to give to charity.might be rendered thus: (5-6a) Pro-attitude (actions that are F); Belief (giving to charity is F) Ridge is arguably right that moral language somehow involves general attitudes of that kind, but is it plausible that we actually express them whenever we make moral judgments?It seems much more plausible to say that in making moral judgments we take such attitudes for granted.Now if it is true that such general attitudes are not expressed by our utterances, then how can we characterize contents of that kind?The current debate on various kinds of secondary contents (such as presuppositions, conventional implicatures, use-conditional items etc.) is confusing in many respects (see, e.g., Horn, 2013;Tonhauser et al., 2013;Potts, 2015;Gutzmann, 2021;Sander, 2022b).I cannot hope to settle that debate in this paper, so I shall generally remain as non-committal as possible.I think, however, that the specific notion of pragmatic presupposition may come in handy here, for two reasons. 21First, it is widely agreed that paradigmatic presupposition triggers are subject to a constraint Tonhauser et al. (2013) have called "strong contextual felicity" (compare Section 3): using items such as "too" is acceptable only if what is encoded by them is mutually known by the interlocutors. 22By the same token, when using loaded words such as "bastard" or "blasphemous", we typically presume that our interlocutors share our general attitudes, although we may express our attitude towards a particular target in a potentially informative way.(For that reason, words often become obsolete once general attitudes change.For instance, there aren't many people who still see "chaste" as a word of theirs.) essentially draws on the assumption that "there are richer and more intensive terms available for unpleasurable affects than for pleasurable ones", and this assumption seems correct: while many languages contain a plethora of pejoratives, amelioratives are quite rare. 27Now Wundt's central idea is that a particular fact about our mental life may account for the rarity of amelioratives and also for the observation that terms of abuse are sometimes used as though they were terms of endearment: It is only the urge to accentuate the feeling as much as possible that brings about this result.Like the richer vocabulary for unpleasurable affects, it is probably based on the characteristic of our emotional life that forms of unpleasure can reach greater degrees of intensity.Wherever a very strong pleasurable affect is to be expressed, a designation that in fact belongs to the realm of unpleasure is easily substituted for.(Wundt, 1922, p. 578, my translation) If Wundt is right, then there is a fundamental asymmetry between pejoratives and amelioratives: pejoratives can be used to express positive attitudes, but not vice versa.So the contribution of pejoratives is more flexible (and therefore more elusive) than that of their positive counterparts.
This high degree of flexibility means of course that it is very hard to come up with an analysis of pejoratives that can account for all of their vagaries.So the analysis I have presented here is, admittedly, incomplete, but I still think it is well poised to handle the standard use of terms such as "jerk" or "bastard".

ACKNOWLEDGMENTS
For feedback on earlier drafts of this paper, I would like to thank Laurence R. 2 There is an abundance of such words.In what follows, I will use the comparatively non-nasty word "jerk" and its slightly more aggressive cousin "bastard" as standard examples.
Horn and anonymous reviewers for this and other journals.Work on this paper was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) -439302327.Open Access funding enabled and organized by Projekt DEAL.ORCID Thorsten Sander https://orcid.org/0000-0002-0880-8843ENDNOTES 1 I will often use that sentence (or variants thereof) as an example.Since quod licet Iovi non licet bovi, I will change the name in what follows. 12 Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ejop.13007 by Thorsten Sander -Universitaet Duisburg-Essen , Wiley Online Library on [20/08/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License does not carry the descriptive claim that the addressee is a cardinal.It is, rather, a means of politely addressing a cardinal.Now is there a principled means of distinguishing descriptive felicity conditions from descriptive contents Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ejop.13007 by Thorsten Sander -Universitaet Duisburg-Essen , Wiley Online Library on [20/08/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ejop.13007 by Thorsten Sander -Universitaet Duisburg-Essen , Wiley Online Library on [20/08/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ejop.13007 by Thorsten Sander -Universitaet Duisburg-Essen , Wiley Online Library on [20/08/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License attitude towards persons that are F.As stressed before, these requirements are often quite elusive, but nevertheless