Dogwhistling as a narrative-evoking form of communication

Abstract

In this essay I defend the view that dogwhistling is a a speech act performed with a narrative-evoking perlocutionary effect in the so-called target audience. What is evoked is a certain kind of narrative, previously endorsed by the relevant audience, which endows its members with the use of some linguistic expressions (and some non-linguistic representations) with non-conventional, derived meanings. In the dogwhistling scenarios, those derived meanings are recovered and put to work by means of different mechanisms, which has an impact on the emotional and practical attitudes of the target audience. The covert message is thus inferred as the product of the recovered meanings at work and their emotional and practical impacts on the audience in the new contexts of use, which determines a new pragmatic meaning dimension for the expressions in play. Although the phenomenon has been frequently analyzed in connection with examples of political discourse, it is common to cinematographic and literary intertextual references, and, more generally, to all those occasions in which communication relies on the narrative dependance of linguistic use.

Keywords:
Dogwhistling; Intertextuality; Narrativity; Perlocutionary effect; Narrative-evoking effect

Dogwhistling as a narrative-evoking form of communication

Dogwhistling involves a covert form of communication; more specifically, it is a speech act in which two different messages, an open message and a covert one, are conveyed to two different audiences, a general audience and a specific one, respectively. The last audience, at which the covert message is directed, is usually characterized as the target of the dogwhistle. The main examples have been usually drawn from political discourse, which makes for the perfect arena for the delivery of this kind of “double message” -given that politicians need to address and engage different groups with diverse interests and backgrounds. Witten (2008), Saul (2018) and Lo Guercio and Caso (2022) maintain that dogwhistles are specializations of general covert messaging. I will be defending a related but different kind of hypothesis: dogwhistles are a specific variety of the general phenomenon of narrative sensitivity, namely, the fact that the narrative frameworks in which linguistic expressions are used can endow them with evocative power. According to this view, dogwhistling basically involves evoking those narratives in a context in which they do not prevail; the evoked material is the background against which different mechanisms may be in play for the inference of the covert message.

A clarification point is in order. Saul (2018) has distinguished between overt and covert dogwhistles on the basis of the kind of effects they achieve on the target audience: on the one hand, overt dogwhistles are performed with an intention that is to be recognized by the audience and, hence, the effect they produce in the audience is taken to depend on their recognition of that intention; covert dogwhistles, on the other hand, are performed with an intention that is not to be recognized by the audience (since its recognition would turn it ineffective), and, hence, their effects are considered to be perlocutionary ones. In this essay I will focus on overt dogwhistles, namely, speech acts performed with an intention that is to be recognized by the target audience; however, I will defend the view that, on top of that, they are speech acts characterized by producing a particular kind of perlocutionary effect, a narrative-evoking one, namely, the recalling of a certain narrative, and that it is only against that background that the corresponding illocutionary intention can be recognized. Therefore, from this perspective, whereas all dogwhistles are speech acts with a perlocutionary dimension, the difference between the overt and the covert ones is that only the former are speech acts made with the intention to produce a narrative-evoking perlocutionary effect in the target audience, whereas the latter’s intended perlocutionary effect is not narrative-evoking, namely, it does not consist in the evocation of a certain narrative. As is usual with intentions oriented towards perlocutionary effects, that intention may not be recognized by the target audience; what makes the act at stake an overt dogwhistle is, though, the fact that the intention to convey a concealed message against that narrative background is clearly to be recognized.

The structure of the paper is the following. In section 1, I present some examples that are intended to show how widespread the phenomenon of narrative-evoking communication can be taken to be. In section 2, I introduce a specific proposal concerning dogwhistles, with regard to an example, considered paradigmatic, based on the idiosyncratic use of language internal to a group (a family). In section 3, I apply the proposed analysis to some mostly well-known political examples. Finally, I briefly conclude with a general summary.

1. Motivating the proposal: intertextuality and narrativity

I would like to motivate the analysis by focusing on some phenomena which, from my perspective, involve the same kind of narrative-evoking effect that can be ascribed to the so-called dogwhistles. The examples I will present are cases of intertextuality, in a broad sense of the term in which it is not just a literary phenomenon but one that can encompass different kinds of media. Roughly, intertextuality has been originally characterized as the “dialogue” between different texts by virtue of which some characters, themes, ideas or situations, originally introduced in some texts, can recur in other ones -but then applied across different media, so that ‘text’ in the definition should be understood as embracing any kind of symbolic system.¹ Let’s consider, first, the presence of references to an older movie in a newer one. To pick up a random example, the opening scene of Dress to Kill (1980) by Brian De Palma, featuring a woman who is taking a shower, involves a reference to the famous scene in Alfred Hitchcock’s Psycho (1960), where a woman gets killed while taking a shower. I would say that watching the De Palma’s opening scene will have an evocative effect in those people who have a background knowledge of cinema, as a consequence of which they will be able to classify the movie within a certain genre, namely, suspense or psychological thriller, a sub-genre of thriller in which the plot revolves around the deranged psychological profile of some characters, outstandingly represented by Psycho. That effect depends on those spectators belonging in or, as I prefer to put it, endorsing (to a certain extent or in a certain way) what can be described as a narrative framework, within which certain principles and categories are put to work for the analysis of film, namely, a cinematographic criticism narrative.² As is clear, there are different levels of participants in that narrative framework, ranging from the strong participation of experts to the weaker participation of those who simply enjoy movies and read critical reviews on newspapers and magazines. Whatever level of participation they belong in, they can be taken to constitute the director’s target audience for the film.

The last remark uncovers a closely related but different imprecision: it is not only that the expert can be taken to know more about film criticism than the curious reader but also that the last one may be considered not to know anything at all -she may be just familiar with some categories or accept them uncritically as useful tools for analyzing movies. In general, the issue is that endorsing a certain narrative framework encompasses having different kinds of epistemic relations to its assumptions. As it will become clearer across the examples to follow, depending on the case, it may range from merely being familiar with some assumptions to being committed to the truth of some claims and the adequacy/propriety of some norms and acting on their basis (as in the political examples to be seen in section 3).

So, more specifically, when watching the opening scene of Dress to Kill, the spectator who is at least familiar with some cinematographic principles and categories and their main examples will be able to infer that the woman at stake is in the same situation as the woman in Psycho’s famous scene, namely, that she is completely oblivious of the serious danger she is in and that they should feel (fictionally) worried about her.³ Moreover, there is an emotional and/or practical side that is part of the evocative effect: a particular way of being predisposed to enjoy that genre of movie, a psychological thriller. The narrative-evoking effect allows thus for a different (cognitive and emotional) interpretation of the opening shower scene of the De Palma’s film: in other words, the scene can be considered to involve a complex concealed message that can only be accessed by those spectators who belong in the cinematographic criticism narrative. On the other hand, the cinematographic-criticism outsiders are not in a position to anticipate the murdering of the woman in the shower, they will not immediately realized that the movie in play is a psychological thriller, and they will then not be emotionally predisposed to enjoy a movie of that genre.

A clarification point may be useful: there is a sense in which the movie has a broader audience as its target, namely, not just the cinematographic-criticism informed viewers. When I say that the target audience can be reduced to them, I have in mind a sense of ‘target audience’ according to which it is constituted by those people who can have a different, less straightforward, more elaborate and nuanced interpretation of the film.⁴ They are the target audience on the assumption that artists aim at the fullest interpretation of their works; anyway, I concede that they could be considered to be not the target audience but a target audience, among many different ones (namely, taking into account that some artists may be totally indifferent about how their works are interpreted). Notice, though, that I am not making a point about the superior quality of the informed viewers’ aesthetic experience -I am just saying that their interpretation is more elaborate and nuanced than the one delivered by the general audience.

The same phenomenon can be exemplified by appealing to those films that contain references to literary works, such as, again, to take a random example, Strategia del ragno (1970), by Bernardo Bertolucci, based on Jorge Luis Borges’ short story “Tema del traidor y del héroe” (1944). Before expanding on the example, I should point out that it revolves around not a particular scene but the general plot of the story. Borges’ short story narrates an event that (fictionally) took place in Ireland in 1824: the leader of a group of conspirators, Fergus Kilpatrick, urges his main subordinate, James Nolan, to find and execute the person who, he knows for sure, is going to betray their imminent rebellion. As it happens, that person turns out to be Kilpatrick himself, who then asks Nolan to comply with his order but maintain the betrayal on the cover. Nolan decides to present the execution as if it were an unfair murder in the hands of their enemies, in dramatical circumstances. As indicated by its title, the short story exemplifies what Borges calls the theme of the traitor and the hero, namely, a story featuring a character who appears to be good while really being bad, according to which it turns out to be better for most people not to know the truth and abide by the appearances. The theme reappears in the above-mentioned film, Strategia del ragno: now the action takes place in Tara, a fictional town located in Italy, in 1936.⁵ The movie tells the story of a young man, Athos Magnani, who goes to Tara to find out about his father’s death. The father, after whom he was named, is remembered by everyone in town as a brave combatant and an unfair victim of fascism. As it happens, he was the leader of a partisan group who had planned to kill Mussolini but betrayed his friends by telling the plan to the fascists. With his acquiescence, the friends had decided to kill him for treason but, at the same time, keep that in secret and tell everybody in town that he had died defending the partisans’ ideal. His death took place in full sight, at a theater, during the performance of Rigoletto. Thereafter, he was remembered as a hero by all the inhabitants of Tara.

Now, I would say that the film has an evocative power for those spectators with some background knowledge of literary criticism:⁶ in watching it, they can evoke the general view that there are certain themes that may recur in different fictional narratives (even if they are of different kinds or involve different media), and that, among those themes, there is the one of the traitor and hero, introduced and paradigmatically exemplified by Borges’s short story “Tema del traidor y del héroe”. Those informed spectators can be described as endorsing a literary criticism narrative that gets evoked in watching the film. Again, we may wonder how much knowledge of literary criticism, if any at all, is required to be considered as endorsing such an informed narrative framework. My answer is, as before, that there are different kinds of endorsement, from the characteristic expert to the curious layman. The informed viewers, whatever level of information they take part in and whatever epistemic attitude, within a certain spectrum, they have concerning that information (doubt, provisional acceptance, belief, knowledge, etc.), are to be considered the director’s target audience for the film -with all the above-mentioned provisos.

That narrative-evoking effect, namely, the recollection of a literary criticism narrative, with its provision of the film’s theme, enables the target audience to access an interpretive dimension that cannot be accessed by the general audience: for instance, only the informed viewer will be initially suspicious about the many heroic images of the late Athos Magnani, will notice that the character of the woman (Draifa, who has no parallel in the short story) is hiding something about his real moral profile (in spite of her being still in love with him), and will notice that the friends are artificially laudatory of Magnani’s character. As for the emotional and/or practical impact, from the very beginning, only in their case, the expectations will conform to the final discovery of a dual, disappointing and obscure figure. These are all aspects of the director’s covert message, articulated by his use of the literary theme.

Finally, to include an example of an intertextual reference in a literary text, consider “Adam Cast Forth”, the poem by Borges replicated below:

(1)

Was there a Garden or was the Garden a dream?

Amid the fleeting light, I have slowed myself and queried,

Almost for consolation, if the bygone period

Over which this Adam, wretched now, once reigned supreme,

Might not have been just a magical illusion

Of that God I dreamed. Already it's imprecise

In my memory, the clear Paradise,

But I know it exists, in flower and profusion,

Although not for me. My punishment for life

Is the stubborn earth with the incestuous strife

Of Cains and Abels and their brood; I await no pardon.

Yet, it's much to have loved, to have known true joy,

To have had - if only for just one day -

The experience of touching the living Garden.⁷

As is clear, the highlighted words can be interpreted as alluding to the biblical narration, which can be construed as a religious narrative encompassing, among many others, the following assumptions supported by the Genesis: God created the first human beings, Adam and Eve, and make them live in a paradise of perfect happiness called Eden; God has definitively expelled Adam and Eve from it, due to their being disobedient and arrogant; the divine punishment, consisting of a life subject to evil, war and pain, applies to all their descendants, starting with their children, Cain and Abel; etc. Notice that the relevant narrative is an independently established religious narrative, alluded to by the author, that some readers (maybe most of them) are familiar with. I would then say that reading the poem has a religious narrative-evoking effect in those readers, who function as the author’s target audience. In this case, it is clear that endorsing the narrative only requires that the person should be familiar with it, independently of whether they are committed to the truth of its main claims or not. Due to that endorsement, the target audience is in a position to understand that the garden mentioned in the poem is Eden, the paradise created by God for Adam and Eve. Moreover, for them, words like ‘Adam’, ‘Cain’, ‘Abel’, etc. are names for biblical characters they know about. The narrative-evoking effect allows for accessing the message concealed in the form of a general metaphorical meaning: going holistic, the poem expresses the author/narrator’s certainty about the reality of human happiness, which seems, though, brief and ephemeral, namely, mostly as distant from our present as the biblical paradise (and definitely distant from the author/narrator’s, immersed in sadness, as indicated by the “fleeting light” and Adam’s “wretched” present).⁸ Besides, the narrative-evoking effect comprises an emotional impact on the target audience: from the beginning, they will be predisposed to deal with an existential poem, focused on the issue of whether happiness is real or just a dream, which most probably will put them in a reflective mood. In contrast, the readers who are unfamiliar with the religious narrative of the Bible, namely, the outsiders, may find it hard to understand why the story of a character named Adam may have an impact on the issue of happiness, or why happiness is supposed to be related to the occasional sight of a certain garden.

To summarize my main point, all the examples presented in this section have something in common: when faced with a visual scene/a story plot/a family of words dominating a poem, those people endorsing a certain related narrative will experience a narrative-evoking effect. That effect allows for a different, more elaborate and nuanced interpretation of the scene/story plot/poem, both cognitively and emotionally. In other words, the derived meaning (in a broad sense of the word) that can be assigned to the scene/story plot/poem against the background knowledge of an evoked narrative can be taken to constitute a concealed or covert message, intended only for a specific audience, the target one. Now, as I anticipated in the introduction, the usual cases of dogwhistling described in the current philosophical and scientific reflection can be considered to be examples of this more general phenomenon, and hence might be analyzed along the same lines. I turn to develop the analysis with some more detail in the following section.

2. The notion of a narrative-evoking perlocutionary effect: a model example

In refining the analysis, I will focus on another literary example, taken from Natalia Ginzburg’s novel Lessico famigliare (1963, translated into English as Family Lexicon). In this book, the author narrates a fictional story based on her life, starting from childhood, during which she lived with her parents and four siblings. The general background is Italy, during the rise of fascism, to which the Levis (Natalia’s maiden name) were strong opponents. What I would like to emphasize is, though, another aspect that is meticulously thematized by the author, namely, the idiosyncratic use of language that can be an identifying feature of a family. Her examples include different kinds of expressions: ‘a Barbison’, ‘most eminent Signor Lipmann’, ‘white lady cutlet’, “Don’t say it’s the teeth", “That girl’s going to marry the gasman”, “I cannot go on painting”, “Sulfuric acid stinks of fart”, "You too have your little things”, “The Brot shot in the pot”, “We haven’t come to Bergamo for campaigning”. In Ginzburg’s own terms:

hose phrases are our Latin, the dictionary of our past, they’re like Egyptian or Assyro-Babylonian hieroglyphics, evidence of a vital core that has ceased to exist but that lives on in its texts, saved from the fury of the waters, the corrosion of time. Those phrases are the basis of our family unity and will persist as long as we are in the world, recreated and revived in disparate places on the earth whenever one of us says, “Most eminent Signor Lipmann,” and we immediately hear my father’s impatient voice ringing in our ears: “Enough of that story! I’ve heard it far too many times already!” (Ginzburg 1963: 20; my translation)

And she adds: “If my siblings and I were to find ourselves in a dark cave or among millions of people, just one of those phrases or words would immediately allow us to recognize each other” (Ginzburg 1963: 20; my translation). The phenomenon pointed out by Ginzburg is, in my view, what lies at the core of dogwhistling. To explain this claim, I will focus on one of the sentences offered in the novel:

(2) I don’t recognize my Germany anymore!⁹

This sentence used to be uttered by Natalia’s mother in the distant past, when parents and children lived all together in the family home. Its use was not literal but metaphorical: the mother used it to express surprise and distress about the fact that something had drastically changed -by alluding to a common acquaintance who had used those words in reference to Germany’s changes after the war. Notice that the meaning (in a generic sense of the word) of the mother’s speech act had the usual two components: first, a propositional content that is metaphorical, and, secondly, an illocutionary force that is expressive.¹⁰ I would say that the expressive use of that metaphor, internal to the family, is part of a private narrative, namely, a framework in which some expressions can be systematically used with non-conventional, derived meanings by virtue of the shared life experiences and practices of a certain group. In that sense, a family can be considered the source of a special (non-conventional) descriptive and expressive aptitude. Now, as told in the book, let’s imagine that the same sentence is used on the occasion of a present gathering of the siblings, in which they can’t help remembering their childhood. Their use of (2) in the new context seems to have an unavoidable evocative power for them, in the sense that it will be strongly associated with the previous use. My main point is then that the present use of the sentence causes, in the family members, what I have been calling a narrative-evoking effect: they evoke the memory of the family narrative, in terms of which they ascribe a certain metaphorical meaning to the sentence. Memory takes them back “home”, and home constitutes the source of a private narrative, namely, expressions get inserted in a framework that crucially encompasses some private and idiosyncratic linguistic uses. In general, a family narrative can be taken to include personal terminologies and definitions, as well as some idiosyncratic use of the lexicon and all kinds of non-figurative uses of sentences, inspired by their members’ common share of experiences, practices and life circumstances.

In more technical terms, I would say that the use of that sentence in the context of the siblings’ gathering is a speech act with a perlocutionary dimension, namely, a speech act that has the effect of evoking, or recalling one another, the memory of their family narrative. It is important to take into account that evoking a memory is not a kind of illocutionary act but a perlocutionary one, namely, it causes an effect in an audience that is not achieved by means of a Gricean mechanism, that is, the recognition by the audience of the speaker’s intention: in hearing the sentence at present any family member remembers the family narrative not by recognizing the speaker’s intention to cause them that memory but just because the sentence played a role in their past linguistic practices. As originally suggested by Strawson (1964), an illocutionary effect can be conceived as one that is achieved by means of a Gricean mechanism -a hearer is warned about something because they recognize the speaker’s intention to warn them about it-, whereas a perlocutionary effect is one that is not achieved in that way -a hearer does not get impressed because they recognize the speaker’s intention to impress them, if there is any-. From my perspective, evoking the memory of a private narrative framework in a friend or a sibling by uttering a certain sentence (or, for that matter, humming a certain melody or showing them a certain picture) is not like warning but rather like impressing someone. In the impressing case, a hearer gets impressed not because they recognize the speaker’s intention to impress them but because they like what is said or how it is said (or because the hearer feels respect or admiration for the speaker). Likewise, in the case at hand, the hearer evokes the shared idiosyncratic narrative not because they recognize the speaker’s intention to cause them that memory but because they belong in the same family and, as a consequence, they took part in the same life experiences and practices. In a nutshell, impressing an audience requires them to be predisposed in a certain way, for instance, that they show interest in the subject, that they get surprised or motivated by the topic or the approach, that they feel respect or admiration for the speaker; likewise, the cases of evoking the memory of a narrative in an audience like the ones described in Family Lexicon can be considered to require them to belong in the same group and hence have had some common life experiences and practices. I would then say that having shared some life experiences and practices is what makes the narrative-evoking effect possible in the case at stake -as much as being interested in the topic, respecting the speaker, being motivated by their approach, etc. is what makes the impression effect possible.

The evoked family narrative might be loosely summarized as follows: their mother systematically uttered that sentence anytime she was negatively affected by a change; her utterances were inspired by the memory of a German friend of theirs; she (the mother) was prone to that kind of expressive uses because of her literary character; her husband, their father, being a man of science, despised that way of talking; etc. The family members can be considered to endorse that private narrative. Now, on the one hand, the evocative effect is what allows for the metaphorical interpretation of the utterance that was characteristic of the previous use; as mentioned before, (2) was systematically used to express worry and distress about any circumstance involving a drastic change. This meaning (in the generic sense before mentioned), assumed in the context of their present encounter, is derived from the context of the past family conversations, that is, one that was clearly dependent of the above-mentioned family narrative. Loosely alluding to Stalnaker (1978), it may be thought that the original context -let’s call it the memory context- contains a set of propositions that, strictly speaking, are not part of the present context -let’s call it the present utterance context-. The last one includes a time where the parents, being both dead, are prone neither to use the sentence nor to react to its use in any particular way; anyway, it gets completely permeated by the time of the memory context. That is why I would say that the two contexts inevitably overlap. To use Stalnaker’s terms, by virtue of the remembrance, the common ground of the current conversation gets pervaded by the narrative assumptions held in the past.

On the other hand, the evocative effect seems to have an emotional and/or practical component, since the recalling is a subjective experience, something that is felt by the family members in a certain way. The recollection of the family narrative not only allows for the recovery of an old meaning but also has an emotional effect in the hearer: the siblings not only interpret the present use in accordance with their recollection of the narrative but they are also emotionally affected by that recollection, in a way that may impact on their present decisions and actions. (This will be specially relevant regarding the political examples that will be considered in the next section.) In connection with this point, it should be taken into account that it has been suggested that the Stalnakerian common ground can be thought to encompass not just propositions but other kinds of entities as well, such as to-do-lists (Portner 2004) and norms concerning the propriety of having certain feelings or emotions (García Carpintero and Marques 2020, Marques 2022). In the current scenario, I would suggest, without developing the idea any further, that the common ground gets updated not just with the content of the narrative assumptions specified above but also with a norm concerning the propriety of being nostalgic and feeling specially close to one another when faced with the utterance of a (narratively-loaded) sentence like (2).

The evocative effect can be considered to determine a pragmatic dimension of the sentence at stake: upon hearing the new utterance the family members evoke the memory of a family narrative that makes an old metaphoric use available for them. Moreover, the evocative effect can be partly conceived of as constitutive of a pragmatic expressive dimension of the new use, since it contributes a further expressive dimension (that adds to any previous expressive dimension the old use may have had).¹¹ To put it differently, on top of expressing whatever that sentence was metaphorically used to express in the past (worry and distress about a drastic change), for the siblings it now expresses something else as well, most probably, some nostalgia about their lost childhood and some feeling of special closeness, of being “bound” by their common past. These are the two aspects of the covert message that is inferred by any member of the family, upon hearing an utterance of (2) in the context of their adult encounters.

At this point, it is convenient to make some distinctions concerning the audience. It is clear that an utterance of (2) will not have any narrative-evoking effect for someone who does not belong in the Ginzburg’s family, namely, the target audience. As a consequence, the family outsiders will only understand its conventional meaning -namely, the speaker’s expression of distress for not recognizing their own country, Germany. Outsiders are not in a position to recover any derived meaning since they did not take part in the life experiences and practices that gave rise to the idiosyncratic linguistic uses of the family. However, it may be thought that the possible audience encompasses, on top of the target audience and the outsiders, a third category, constituted by those people who can access the derived meaning without having taken part in the underlying uses, just for having heard about them from the outside. That category may be thought to comprise, in the particular example at stake, a family friend who heard about the existence of their idiosyncratic uses without taking part in them. In other terms, the people who did not take part in the family uses can be classified in two groups: the full-blooded outsiders, namely, those who had no clue about them, and the peripheral outsiders, those who had just heard about them. So, it is important to note that, in contrast with the examples presented in section 1, there are examples in which knowing about/being familiar with the narrative does not count as endorsing it. If that is so, those who have a weaker epistemic relation to the corresponding narrative might be taken to form an intermediate category between the target audience and the general audience.¹²

There is an important aspect to be pointed out before moving on. Notice that a narrative-evoking speech act may involve just one (kind of) audience-in other words, its target audience might be the only audience in play. In that case, there will be just one message involved, retrievable by virtue of the whole audience’s evocative capacity -hence no distinction between two kinds of messages, an open and a covert one, can be made. As an example, imagine a conversation between two old friends in which one of them utters a sentence belonging in a film that they both cheered when they watched it together during their adolescence: the utterance can be taken to evoke in the interlocutor the memory not just of the film but also of the old times spent together, in the far-away past. In my opinion, this makes for an example of a narrative-evoking speech act that is not a dogwhistle, which concords with my initial hypothesis according to which dogwhistling is a variety of a more general, narrative-evoking form of communication.¹³

3. Applying the analysis to some political examples

I will start by introducing a rather simple example, taken from Argentina’s political history, in which the covert message is triggered by the use of a single word, whose derived meaning is fixed by what can be called a political narrative:

(3) People who, like me, live only for the pueblo, need that solidarity. Therefore, whenever I have spoken to the pueblo, more than orders, I have given advice. A president who advises, more than a president, is a friend, and that is precisely what I want to be of my pueblo: a friend. Always fulfilling the first truth established in our Peronist catechism, which says that true democracy consists in the government doing only what the pueblo want and defending only one interest: that of the pueblo.¹⁴

Here, the crucial word responsible for the evocative effect is ‘pueblo’, and the evoked political narrative in whose framework the word gets a meaning different from its conventional one is explicitly mentioned by means of the expression ‘Peronist catechism’ (catecismo peronista), which can be loosely summarized as follows: the pueblo is the working class; the working class has been oppressed by the traditional political class belonging to the Conservative party; the Peronist party is the only one that has done something for the working class, since it strongly supported the CGT (the association that gathers all the main unions in Argentina); only the Peronist party can represent and defend the interests of the working class; etc. As a consequence, the target audience, namely, those people endorsing the Peronist political narrative, will recover the non-conventional meaning of ‘pueblo’ and then infer the covert message to the effect that the speaker is addressing the working class, of which he presents himself as a friend (and, probably, as a member of the group), and that he commits himself to the claim that a true democracy consists in the government representing only the working class. Moreover, the evocative effect also produces an emotional impact (the target audience will feel enthusiastic, moved, and other feelings of the likes) and a determination to act based on their loyalty to the evoked doctrine (they will be prone to vote for their leader and similar dispositions).¹⁵ Both constitute the emotional and/or practical component, as opposed to the previously mentioned cognitive component, of the covert message.

As must be clear, endorsing the Peronist political narrative amounts to being committed to the truth of its main assumptions and being disposed to act on the basis of that commitment: this is what it takes to belong in the target audience. Those people who are merely familiar with the political narrative at stake (namely, who have a weaker epistemic relation and no practical commitment) can be considered, in this case, to be peripheral outsiders, who won’t experience the full-blown evocative effect: even if they can retrieve the non-conventional meaning of ‘pueblo’, they will not get the emotional and practical aspects of the covert message (since they will tend to have some misgivings about the speaker’s sincerity and be prone to vote against him). Finally, notice that, taking into account the context in which the speech was delivered, namely, Buenos Aires central square in 1953, there are not full-blooded outsiders -those would have been foreigners without any knowledge of the political history of Argentina, for whom the message could only have conveyed the leader’s devotion for the whole population of the country.¹⁶

By virtue of the dogwhistle hereby exemplified, ‘pueblo’ has got a pragmatic meaning dimension (making it synonymous with ‘the working class’) that is widely spread in the Argentine community. In contrast with the Ginzburg example, the evoked narrative on which it depends, namely, the Peronist catechism, is not private but socially shared. But the core phenomenon in play is the same in both cases: the use of an expression, strongly associated with a particular narrative, causes an evocative effect in the audience endorsing that narrative, allowing thereby for the recovery of a non-conventional meaning and the ensuing emotional and practical impact (namely, the inference of the covert message). For them, any context of utterance gets updated with the assumptions and norms that were in force in the memory context, namely, that in which the narrative prevailed.

A slightly more complex example is the one, largely quoted in the philosophical and political science reflection on dogwhistles (Khoo 2017, Saul 2018, Henderson and McCready 2019, Lo Guercio and Caso 2022), featuring a sentence included in the speech given by George W. Bush during the campaign for the 2004 presidential election in the USA:

(4) Yet there’s power, wonder-working power, in the goodness and idealism and faith of the American people.¹⁷

Here, the evocative effect is also triggered by the use of a single, though more complex, expression, ‘wonder-working power’. It evokes the memory of a religious narrative, characteristic of the Evangelical community, whose central claims include something along the following lines: it is on the basis of faith and repentance from sin that believers are made right of their transgressions of the law of God rather than on the basis of their good works (justification by faith or conversionism); human writers and canonizers of the Bible were led by God with the result that their writings can be designated the word of God (biblical authority or biblical inspiration); atonement, the saving death and the resurrection of Jesus, offers forgiveness of sins and new life (crucicentrism); the gospel has to be shared in diverse ways that include preaching and social action (activism); etc. As attested by the last claim, the narrative crucially enforces the followers´ compliance with certain practices, including singing praise and worship hymns, one of which contains the expression ‘wonder-working power’. So, Bush’s target audience in uttering (4) is a certain community among the USA population of possible voters, namely, the community endorsing that particular religious narrative. Notice that in this case endorsing the religious narrative requires as well that the person should take its main assumptions to be true (and its main commands to be normatively appropriate).

According to the analysis I am putting forward, Bush’s speech act has thus a narrative-evoking effect on his target audience: it evokes the religious narrative they actively endorse, in whose framework the expression ‘wonder-working power’ is commonly used as part of a worship song. But, whereas in the Perón’s example the reinterpretation of a single expression (by recovering its non-conventional meaning) was enough to infer the cognitive part of the covert message, in this case it also requires a reinterpretation of the whole utterance in which the expression appears. More specifically, the mechanism involved in getting an additional, pragmatic meaning for the whole utterance can be taken to be a conversational implicature, to be reconstructed along the following lines:

1. the speaker, namely Georges W. Bush, has used the expression ‘wonder-working power’;

2. that expression features in the lyrics of one of our worship songs;

3. only those people who endorse the Evangelical principles and practices are familiar with the lyrics of that hymn (in other terms, if someone is familiar with the lyrics of the hymn, then she endorses the Evangelical principles and practices, since singing that hymn is a practice internal to the Evangelical community);

4. the speaker knows that (ii) and (iii) and he knows that we know that (ii) and (iii);

5. the speaker has done nothing to prevent that we believe that he endorses the Evangelical principles and practices;

6. the speaker endorses the Evangelical principles and practices.

The implicature seems to follow from the assumption that the speaker is respecting the Maxim of Mode: if he openly quotes the words of the hymn in the context of a serious (non-comical or ironical) assertion and is collaborative, he must be in part alluding to what they mean in the context of the hymn, namely, the power of JesusChrist as conceived and celebrated by the Evangelicals. As it happens to many pieces of political discourse, the content of the conclusion to be inferred is, though, vague: it is not clear what kind of endorsement is at stake -maybe Bush is committed to the truth of the Evangelical principles and the moral duty of complying with their practices, just like the Evangelicals themselves, or maybe he is familiar with them and very sympathetic but not committed to those principles and duties in the same way.¹⁸ Moreover, different members of the community may come to identify a different message according to their previous beliefs and expectations about Bush -for instance, if they know he used to be a member of the Episcopal Church, they may be a bit suspicious about the strength of his commitment, but if they ignore that fact they may be, instead, more confident about it. Notice that vagueness is something the speaker, that is, Bush himself, is counting on, since it clearly serves his political purposes: he wants those people who are in turn familiar with the religious narrative but not committed to it (the peripheral outsiders) not to infer with certainty that he has a strong commitment to the Evangelical doctrine, since he is of course also interested in getting their votes... Finally, the covert message will have an emotional impact on the target audience: they may feel empathically identified with Bush as a person, enthusiastic about the prospect of his becoming president, and prone to vote for him in the upcoming election.

Another fragment of Bush’s political speech, also profusely quoted by many philosophers (Khoo 2017, Saul 2018, Henderson and McCready 2019, Lo Guercio and Caso 2022), may be analyzed along similar lines. In the presidential debate held during the same political campaign, when asked about what kind of judge he would appoint for the Supreme Court (if there were a vacancy), he replied:

(5) I would pick somebody who would not allow their personal opinion to get in the way of the law. I would pick somebody who would strictly interpret the Constitution of the United States.(...) Another example is the Dred Scott case, which is where judges, years ago, said that the Constitution allowed slavery because of personal property rights. That’s a personal opinion. That’s not what the Constitution says.¹⁹

The use of the expression ‘personal rights’ in the context of the allusion to the Dred Scott case evokes in Bush’s target audience, namely, the supporters of the Republican party, an anti-abortion narrative according to which: abortion is against the law of God; pregnant women have no personal right to decide to interrupt their pregnancy; the Dred Scott case, in which the personal right to property was alluded to in defending slavery, has been usually mentioned in connection with the discussion of abortion; etc. Independently of how the narrative is specified in detail, those people endorsing it (in this case, as in the previous one, those who are committed to the truth of its assumptions and ready to act on their basis) will be in the position to infer a covert message to the effect that Bush is comparing the Dred Scott’s case, in which slavery was defended on behalf of a personal right (to property), to the defense of abortion on behalf of a personal right (to decide not to have a child) and equally opposing both arguments. The mechanism in play seems to be an argument from analogy, which could be reconstructed along the following lines:

1. the speaker, namely, Georges W. Bush, has expressed his commitment to respect the Constitution;

2. the speaker pointed out that the Dred Scott case, which is where judges, years ago, said that the Constitution allowed slavery because of personal property rights, is not based on the Constitution;

3. abortion supporters also invokes human rights in their defense of abortion;

4. the speaker is comparing the invocation of human rights in the Dred Scott case to the invocation of human rights by abortion supporters;

5. the speaker considers that the invocation of human rights by abortion supporters is not based on the Constitution;

6. the speaker considers that there cannot be a constitutional law for abortion and will be opposing it.

Moreover, part of the covert message is getting the target audience’s approval, prompting their enthusiasm about the prospective government policy, and making them vote for the speaker. Notice that those who are merely familiar with the anti-abortion narrative will count, as in the Perón’s example, as peripheral outsiders: they may infer the cognitive part of the covert message but not the emotional and/or practical part, since, far from feeling any enthusiasm for the anticipated policy, they may feel angry and be prone to vote against Bush. On the other hand, full-blooded outsiders, namely, those who completely ignore the anti-abortion narrative will interpret Bush’s remark as a claim about the heinous character of the Dred Scott verdict -the kind of obvious and inconsequential claim that is characteristic of most political speeches.

4. Conclusion

In this essay I have defended the claim that dogwhistling is a a speech act performed with a narrative-evoking perlocutionary effect in the so-called target audience. What is evoked is a certain kind of narrative, previously endorsed by the relevant audience, which endows its members with the use of some linguistic expressions (and some non-linguistic representations) with non-conventional, derived meanings. In the dogwhistling scenarios, those derived meanings are recovered and put to work by means of different mechanisms, which has an impact on the emotional and practical attitudes of the target audience. The covert message is thus inferred as the product of the recovered meanings at work and their emotional and practical impacts on the audience in the new contexts of use, which determines a new pragmatic meaning dimension for the expressions in play. Although the phenomenon has been frequently analyzed in connection with examples of political discourse, it is common to cinematographic and literary intertextual references, and, more generally, to all those occasions in which communication relies on the narrative dependance of linguistic use.

Acknowledgments

The Consejo Nacional de Investigaciones Científicas y Técnicas of Argentina (CONICET) has provided the main funds for the elaboration of this article. Moreover, it has been written in the framework of two research projects, one granted by the Agencia Nacional de Promoción Científica y Tecnológica (PICT 2018-02843) and the other by the University of Buenos Aires (UBACyT 20020170100649BA, Res. 1041/2018). I would like to thank Ramiro Caso and Nicolás Lo Guercio, for having invited me to participate in this special issue, and the two anonymous reviewers for Manuscrito, for their very insightful suggestions.

References

• Austin, John L. 1962. How to Do Things with Words? Oxford: Clarendon Press.
• Bakhtin, Mikhail. 1984. Problems of Dostoevsky's Poetics. Minneapolis: University of Minnesota Press. The original is from 1929 (Leningrad) and 1963 (Moscow).
• Bakhtin, M. 1981. The Dialogic Imagination: Four Essays by M. M. Bakhtin, ed. Michael Holquist, trans. Caryl Emerson and Michael Holquist. Austin and London: University of Texas Press.
• Barthes, Roland. 1957. Mythologies. New York: Hill and Wang, 1972.
• Borges, Jorge Luis. 1944. Theme of the Traitor and Hero. In Ficciones, Part II: Artifices, translated by Anthony Kerrigan. New York: Grove Press, 1962.
• Borges, J. L. 1964. “Adam Cast Forth”. In El oro de los tigres. Buenos Aires: Emece Editores, 1972. Translated into English by Genia Gurarie https://allpoetry.com/Adam-Cast-Forth
  » https://allpoetry.com/Adam-Cast-Forth
• Doggett, T. and Egan, A. 2007. “Wanting things you don’t want: the case for an imaginative analogue of desire”, Philosophers Imprint 7: 1-17.
• Friend, Stacie. 2003. “How I really feel about JFK”. In M. Kieran and D. Lopes (eds.) Imagination, Philosophy and the Arts. London: Routledge.
• García-Carpintero, Manuel and Marques, Teresa. 2020. “Really expressive presuppositions and how to block them”, Grazer Philosophische Studien 97: 138-158.
• Ginzburg, Natalia. 1963. Lessico famigliare. Roma: Einaudi Editore. Translated into English by Jenny McPhee. New York: Review Books Classics, 2017. The references in the text are to the original Italian version.
• Grice, H. Paul. 1957. “Meaning”, Philosophical Review, 66(3): 377-388. Reprinted as ch. 14 of Studies in the way of words (213-233). Cambridge: Harvard University Press.
• Grice, H. P. 1975. Logic and conversation”. In P. Cole & J. Morgan (eds.) Syntax and Semantics 3 (22-40). Cambridge: Academic Press.
• Gutzmann, Daniel. 2015. Use-Conditional Meaning. Studies in Multidimensional Semantics. Oxford: Oxford University Press.
• Henderson, R., & McCready, E. 2019. “Dogwhistles and the at-issue/non-at-issue distinction”. In Gutzmann, D., Turgay, K. (eds.) Secondary content, current research in the semantics/pragmatics interface, vol 37. Leiden: Brill (222-24). https://doi.org/10.1163/9789004393127_010
  » https://doi.org/10.1163/9789004393127_010
• Khoo, J. 2017. “Code words and political discourse”, Philosophical Topics 45(2): 33-64. https://doi.org/10.5840/philtopics201745213
  » https://doi.org/10.5840/philtopics201745213
• Kristeva, Julia. 1967. “Word, dialogue and novel”. In Léon Roudiez (ed.), Desire in Language: A Semiotic Approach to Literature and Art (64-91), translated by Alice Jardine, Thomas Gora and Léon Roudiez. New York: Columbia University Press, London: Basil Blackwell, 1980. First published in Sémeiotikè as "Bakhtine, le mot, le dialogue et le roman" (143-73).
• Lo Guercio, Nicolás and Caso, Ramiro. 2022. “An account of overt intentional dogwhistling”, Synthese 200:203: 1-32. https://doi.org/10.1007/s11229-022-03511-6
  » https://doi.org/10.1007/s11229-022-03511-6
• Marques, Teresa. 2022. “The expression of hate in hate speech”, Journal of Applied Philosophy 10: 1-29.
• Portner, Paul. 2004. “The semantics of imperatives within a theory of clause types”. In Proceedings of Semantics and Linguistic Theory 14, in K. Watanabe and R. B. Young (eds.) (235-252). Ithaca, NY: CLC Publications, Cornell University Linguistics Department.
• Potts, Christopher. 2003. “Expressive content as conventional implicature”. In Makoto K, Shigeto K (eds.) Proceedings of the North East Linguistic Society 33 (303-322). Amherst: GLSA.
• Potts, C. 2005. The Logic of Conventional Implicatures. Oxford: Oxford University Press.
• Predelli, Stefano. 2013. Meaning without Truth. Oxford: Oxford University Press.
• Saul, J. 2018. “Dogwhistles, political manipulation, and philosophy of language”. In Fogal et al. (2018) (360-383). https://doi.org/10.1093/oso/9780198738831.003.0013
  » https://doi.org/10.1093/oso/9780198738831.003.0013
• Searle, John. 1969. Speech Acts. An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.
• Stalnaker, Robert. 1978. Context and Content: Essays on Intentionality in Speech and Thought. Oxford: Oxford University Press.
• Strawson, Peter. 1964. “Intention and convention in speech acts”, The Philosophical Review 73(4): 439-460.
• Walton, Kendall. 1990. Mimesis as Make-Believe: On the Foundations of the Representational Arts. Cambridge: Harvard University Press.
• Witten, Kimberly. 2008. “Dogwhistle politics: the new pitch of an old narrative”. Manuscript originally written and submitted for the Spring 2008 semester at San Francisco State University Courses: Narrative & Discourse Analysis, Professor & Advisor: Dr. Troi Carleton.

• 1
  The concept has been explicitly defined by Kristeva (1967), inspired by the ideas originally put forward by Mikhail Bakhtin in different works, cf. Bakhtin (1984) and the articles contained in Bakhtin (1981). Barthes (1957) is a good example of how the concept has been extended to semiotic studies.
• 2
  The proposed concept of a narrative is similar to what Lo Guercio and Caso call a perspective: “We understand perspectives in very broad terms as sets of beliefs, expectations, non-cognitive attitudes (fears, desires), affective dispositions, practices and habits related to a given issue (politics, religion, art, etc.)” (2022: 10). The only difference is that narratives in the current sense are linguistic entities, whereas perspectives seem to be mixed entities of some sort, more akin to ideologies or worldviews.
• 3
  This analysis is neutral concerning the nature of the attitudes generated in our interaction with fiction, namely, whether the fictional contents at stake are the object of belief, desire, emotion or it is only imagination (i-belief, i-desire, etc.) that is involved. To get a feeling of this debate, see, for instance, Friend (2003) and Doggett & Egan (2007).
• 4
  Witten (2008) offers an example (the use of ‘dirty’ in “Think dirty thoughts”) from the movie for children Looking for Nemo that also exemplifies the present concern. She recognizes two audiences: the general audience constituted by children and a target audience constituted by adults. However, there is a clear sense in which the movie has the children as its main target; but, as she explains, there is another sense in which adults are the main target in as far as they are the target of the covert message. Mutatis mutandi for my example: a director who includes a scene that refers to a famous one can be thought to be conveying a derived meaning based on that reference in the form of a covert message.
• 5
  Tara is a female deity in both Hinduism and Buddhism who personifies compassion and offers salvation from the suffering of rebirth and death. The town name seems to be a symbol for young Athos Magnani’s journey, along which he gets to know the truth about his father.
• 6
  Again, I am conscious that ‘having some background knowledge’ is vague: it may range from having an uncritical familiarity with some literary concepts to being committed to the truth of some critical theories.
• 7
  I have highlighted what I take to be the most relevant words alluding to the biblical narration.
• 8
  Who is to be considered the subject performing the expressive speech act in play? Since it is an issue completely tangential to my present topic, I would like to live it open whether it is the author (namely, Borges himself) or a created character that functions as narrator (based on him, since the poem suggests that he is blind, “Amid the fleeting light, I have slowed myself and queried”).
• 9
  Originally, in Italian: “Non riconosco più la mia Germania!” (Ginzburg 1963: 65).
• 10
  According to Classical speech act theories (Austin 1962, Searle 1969), speech acts have a locutionary dimension (a propositional content) and an illocutionary one (a force). So, in the generic sense of ‘meaning’ alluded to in the text, the meaning of a speech act comprises a proposition and a force.
• 11
  A clarification is needed at this point. A use of a sentence can be expressive in different ways. On the one hand, it can be expressive, as in the example presented in the text, because it has an expressive illocutionary force, namely, because the sentence is used to express a feeling or an emotion. On the other hand, it can be expressive because its propositional content has an expressive dimension determined by the expressive meaning dimension of some of its words (as is the case, for instance, with a sentence containing a slur or any other kind of non-purely descriptive term, such as “That spic/idiot arrived late to the meeting.”) For multidimensional or hybrid semantics, applied to fragments of language that are expressive in the second sense, see, for instance, Potts (2003, 2005), Predelli (2013), Gutzmann (2015).
• 12
  An important clarification point is due. Notice that the Ginzburg example is a literary one. However, in contrast with the examples of intertextuality mentioned in section 1, its literary character is completely irrelevant to the phenomenon under study. The same phenomenon narrated by Ginsburg could be attested in any family’s everyday life. Moreover, we could imagine that Ginsburg’s narration is a factual one, an autobiography in which she narrates an episode of her own childhood (as opposed to a fictional narrative). However, given that it is a fictional narrative, the target audience, the characters of the siblings in their adult lives, is a fictional one (in Walton’s 1990 terms, it belongs in a game of make-believe). An interesting question to ask in this regard is the one about the category in which the readers of Family Lexicon belong. From my perspective, in as far the readers are outside the fictional world, they do not belong in the fictional transposition of any of the above-mentioned categories. However, they can be considered to make for another kind of (fictional) target audience, since, in engaging with the fiction, they can pretend to be family members or, in other terms, they can make as if they were members of the family and actively endorsed the linguistic practices narrated in Ginzburg’s novel (on Walton’s scheme, they would be the target audience in an extended game of make-believe).
• 13
  I am very thankful to a reviewer for Manuscrito whose comment has prompted me to make this point clearer. The example is also hers/his. Notice that this kind of examples are more different from the usual dogwhistles than the examples of intertextuality presented in section 1, which did involve two audiences and the corresponding two messages. As suggested by the same reviewer, a more detailed account of the relation between narrative-evoking speech acts and dogwhistling is due, which I hope to be able to develop on a future occasion.
• 14
  The Spanish original: “Los hombres que, como yo, viven solamente para el pueblo, necesitan de esa solidaridad. Por eso siempre que yo he hablado al pueblo, más que órdenes, he impartido consejos. Un presidente que aconseja, más que presidente es un amigo, y eso es, precisamente, lo que yo quiero ser de mi pueblo: un amigo. Cumpliendo siempre la primera verdad establecida en nuestro catecismo peronista, que dice que la verdadera democracia consiste en que el gobierno haga solamente lo que el pueblo quiere y defienda un solo interés: el del pueblo”. Juan Domingo Perón’s speech, April 15, 1953; my translation. https://www.elhistoriador.com.ar/discurso-de-juan-domingo-peron-en-plaza-de-mayo-15-de-abril-de-1953/
• 15
  One of the characteristics of the covert message, mentioned by different authors (Saul 2018, Lo Guercio and Caso 2022), is that its content is underdetermined.
• 16
  I owe this important clarification to an interesting objection advanced by another reviewer for Manuscrito, who pointed out that introducing a foreign audience would have made sense only if the speech had been made before the UN or some other international organism. In fact, the objection is deeper, since it concludes that the example, even if an instance of a narrative-evoking speech act, cannot be considered a case of dogwhistling. Although I am not completely sure, I tend to disagree, based on the previous considerations: there are still two audiences to distinguish, the target audience and the peripheral outsiders. Whereas the former gets a full-blown narrative-evoking effect, encompassing a message with both a cognitive and an expressive (emotional and/or practical) dimension, the latter only get the cognitive part of the message, since, in as far they are not moved and ready to act on their loyalty to the leader, they miss the corresponding motivational effects.
• 17
  George W. Bush, State of the Union speech, 2003.
• 18
  This example is better than the previous one at showing the underdetermined character of the covert message.
• 19
  George W. Bush, presidential debate, October 8, 2004. https://www.washingtonpost.com/wp-srv/politics/debatereferee/debate_1008.html

• Article info: CDD: 401.

Publication Dates

• Publication in this collection
  30 Oct 2023
• Date of issue
  2023

History

• Received
  20 Apr 2022
• Reviewed
  23 Aug 2023
• Accepted
  29 Aug 2023