1.Introduction

Verbal communication, as is widely acknowledged (e.g., Sperber and Wilson 1995Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell., 2002Sperber, Dan, and Deirdre Wilson 2002 “Pragmatics, Modularity and Mind–Reading.” Mind & Language 17: 3–23. , 2008Sperber, Dan, and Deirdre Wilson 2008 “A Deflationary Account of Metaphors.” In The Cambridge Handbook of Metaphor and Thought, ed. by Raymond Gibbs, 84–105. Cambridge: Cambridge University Press. ; Carston 2002Carston, Robyn 2002 Thoughts and Utterances. The Pragmatics of Explicit Communication. Malden, MA: Blackwell. ; Bach 2007Bach, Kent 2007 “Regression in Pragmatics (and Semantics).” In Pragmatics, ed. by Noel Burton-Roberts, 29–30. Basingstoke: Palgrave Macmillan. ; Wilson and Sperber 2012Wilson, Deirdre, and Dan Sperber 2012 Meaning and Relevance. Cambridge: Cambridge University Press. ; Belleri 2014Belleri, Delia 2014 Semantic Under-Determinacy and Communication. Basingstoke: Palgrave Macmillan. ), involves a certain amount of underspecificity. This is directly related to the fact that utterances abound in contextual variables (such as, e.g., indexical expressions or demonstrative pronouns), vague and conceptually incomplete expressions, ellipsis, structural ambiguities, etc., which must be adequately worked out by the comprehender to get the intended contextual meaning. All this means that what is expressed by utterances is either “too variable or too skimpy to comprise what people mean in uttering them” (Bach 2007Bach, Kent 2007 “Regression in Pragmatics (and Semantics).” In Pragmatics, ed. by Noel Burton-Roberts, 29–30. Basingstoke: Palgrave Macmillan. , 29–30). This idea has been formalized by Carston (2002)Carston, Robyn 2002 Thoughts and Utterances. The Pragmatics of Explicit Communication. Malden, MA: Blackwell. as the linguistic underdeterminacy thesis, which amounts to stating that there is always a gap between what the utterance standing meaning is and what it is used to communicate (see also Belleri 2014Belleri, Delia 2014 Semantic Under-Determinacy and Communication. Basingstoke: Palgrave Macmillan. and the references therein).

All verbal comprehension, then, is taken to involve bridging gaps of various kinds to arrive at the contextually intended meaning. This, as will be argued below, is deliberately exploited in deceptive clickbait headlines. The major of goal of the paper is to examine how headlines of this type work and to show that relevance theory can explain the mechanism at play. It also aims to show that there exists affinity between processing deceptive clickbaits and processing jokes, which is an original contribution of this paper: while there has been research on clickbait headlines before, and comparisons between advertising slogans and headlines have been drawn, the similarity of clickbaits with the punchline effect has not been explored, as far as I know.

The paper is structured as follows. First, the nature of headlines is briefly discussed, and then clickbait headlines and deceptive clickbait headlines are described, with some examples provided to expose their characteristic features. A relevance-theoretic analysis of how deceptive clickbaits function follows. Finally, the idea that there exists affinity between processing deceptive clickbaits and joke processing is discussed. The paper ends with some concluding remarks.

2.Major functions of headlines

Newspaper headlines as we know them today – to be distinguished from headings (serving the purpose of grouping similar news reports) and crossheads (appearing in the body of the text to mark its different sections) – are a relatively recent invention. They were introduced in the role of telegraphic summaries of the news story (Dor 2003Dor, Daniel 2003 “On Newspaper Headlines as Relevance Optimizers.” Journal of Pragmatics 35: 695–721. ) when news material began to be arranged thematically in newspapers, which occurred only in the18th century (Schneider 2000Schneider, Kristina 2000 “Emergence and Development of Headlines in English Newspapers.” In English Media Texts Past and Present: Language and Textual Structure, ed. by Friedrich Ungerer, 45–65. Amsterdam: John Benjamins. ).

Newspaper headlines are markedly different from titles of books, films, paintings, etc. They have been explored from various theoretical and empirical angles, mainly by journalists and linguists (Schneider 2000Schneider, Kristina 2000 “Emergence and Development of Headlines in English Newspapers.” In English Media Texts Past and Present: Language and Textual Structure, ed. by Friedrich Ungerer, 45–65. Amsterdam: John Benjamins. ). Linguistic investigations, with the first extensive monograph on headlines by Straumann (1935)Straumann, Heinrich 1935 Newspaper Headlines. A Study in Linguistic Method. London: Unwin Brothers. published almost a century ago, have focused primarily on the questions of how and what headlines communicate (see Dor 2003Dor, Daniel 2003 “On Newspaper Headlines as Relevance Optimizers.” Journal of Pragmatics 35: 695–721. ). Exploring the former question has led to the identification of a number of syntactic, semantic, stylistic and typographic features, typical of headlinese. In effect, the headline is nowadays recognized as a text type in its own right (Ifantidou 2009Ifantidou, Elly 2009 “Newspaper Headlines and Relevance: Ad Hoc Concepts in Ad Hoc Contexts.” Journal of Pragmatics 41: 699–720. ; Jaki 2014Jaki, Sylvia 2014 Phraseological Substitutions in Newspaper Headlines. Amsterdam: John Benjamins. ).

Seeking answers to the question of what it is that headlines communicate has resulted in identifying their various functions. Stereotypically approached as telegraphic summaries of the text content (van Dijk 1988Van Dijk, Teun 1988 News as Discourse. Hillsdale, NJ: Lawrence Erlbaum Associates.; for a detailed discussion, see Dor 2003Dor, Daniel 2003 “On Newspaper Headlines as Relevance Optimizers.” Journal of Pragmatics 35: 695–721. ), headlines are texts about texts, so they are metatexts par excellence (Iarovici and Amel 1989Iarovici, Edith, and Rodica Amel, R. 1989 “The Strategy of the Headline.” Semiotica 77 (4): 441–459. ). Revealing what the text is about through “the semantic transtextualization of the article” (which has to do with bringing to prominence the core semantic essence of the text), they give the text identity, which suggests that they may be taken to function as its name (Iarovici and Amel 1989Iarovici, Edith, and Rodica Amel, R. 1989 “The Strategy of the Headline.” Semiotica 77 (4): 441–459. , 443). Technically, this contributes to making newspaper page organization easier (Jaki 2014Jaki, Sylvia 2014 Phraseological Substitutions in Newspaper Headlines. Amsterdam: John Benjamins. ).

This kind of informative role that headlines are supposed play is directly related to some other functions. Headlines permit a quick assessment of whether to expect a factual news article or an opinion piece (see Graney 1990Graney, John 1990 “Determination of Fact and Opinion: A Critical Reading Problem.” Journal of Psycholinguist Research 19: 147–166. ). Headlines are also assumed to help readers choose what they want (or do not want) to read (Jaki 2014Jaki, Sylvia 2014 Phraseological Substitutions in Newspaper Headlines. Amsterdam: John Benjamins. , 37), and since they are brief and take little time to process, they enable readers to browse through a large number of items and quickly choose those worth their while (Ecker et al. 2014Ecker, Ullrich, Stephan Lewandowsky, Ee Pin Chang, Rekha Pillai 2014 “The Effects of Subtle Misinformation in News Headlines.” Journal of Experimental Psychology Applied 20 (4): 323–335. ). Furthermore, they provide general context in which to process the text, which facilitates its comprehension and constrains interpretation of content, based on the activation of relevant background knowledge (Bransford and Johnson 1972Bransford, John, and Marcia Johnson 1972 “Contextual Prerequisites for Understanding: Some Investigations of Comprehension and Recall.” Journal of Verbal Learning and Verbal Behavior 11: 717–726. ; Krug et al. 1989Krug, Damon, Byron George, Shawn Hannon, and John Glover 1989 “The Effect of Outlines and Headings on Readers’ Recall of Text.” Contemporary Educational Psychology 14: 111–123. ; Geer and Kahn 1993Geer, John, and Kim Kahn 1993 “Grabbing Attention: An Experimental Investigation of Headlines During Campaigns.” Political Communication 10: 175–91. ; Wiley and Rayner 2000Wiley, Jennifer, and Keith Rayner 2000 “Effects of Titles on the Processing of Text and Lexically Ambiguous Words: Evidence from Eye Movements.” Memory & Cognition 28: 1011–1021. ; Miller et al. 2006Miller, Lisa., Jason Cohen, and Arthur Wingfield 2006 “Contextual Knowledge Reduces Demands on Working Memory during Reading.” Memory & Cognition 34: 1355–1367. ; Ifantidou 2009Ifantidou, Elly 2009 “Newspaper Headlines and Relevance: Ad Hoc Concepts in Ad Hoc Contexts.” Journal of Pragmatics 41: 699–720. ). As Jaki (2014Jaki, Sylvia 2014 Phraseological Substitutions in Newspaper Headlines. Amsterdam: John Benjamins. , 37) emphasizes, they may additionally be a thought-provoking instrument.

The last feature listed is directly connected with another function that headlines play: they are designed to catch attention and arouse interest of potential readers (e.g., Dor 2003Dor, Daniel 2003 “On Newspaper Headlines as Relevance Optimizers.” Journal of Pragmatics 35: 695–721. ; Ifantidou 2009Ifantidou, Elly 2009 “Newspaper Headlines and Relevance: Ad Hoc Concepts in Ad Hoc Contexts.” Journal of Pragmatics 41: 699–720. ; Ecker et al. 2014Ecker, Ullrich, Stephan Lewandowsky, Ee Pin Chang, Rekha Pillai 2014 “The Effects of Subtle Misinformation in News Headlines.” Journal of Experimental Psychology Applied 20 (4): 323–335. ). As is pointed out by some researchers (i.a., Schneider 2000Schneider, Kristina 2000 “Emergence and Development of Headlines in English Newspapers.” In English Media Texts Past and Present: Language and Textual Structure, ed. by Friedrich Ungerer, 45–65. Amsterdam: John Benjamins. ; Ifantidou 2009Ifantidou, Elly 2009 “Newspaper Headlines and Relevance: Ad Hoc Concepts in Ad Hoc Contexts.” Journal of Pragmatics 41: 699–720. ; Jaki 2014Jaki, Sylvia 2014 Phraseological Substitutions in Newspaper Headlines. Amsterdam: John Benjamins. ), there are just two principal functions of headlines that appear superordinate and subsume the other minor ones mentioned above, namely, to inform and to awaken curiosity. Iarovici and Amel (1989)Iarovici, Edith, and Rodica Amel, R. 1989 “The Strategy of the Headline.” Semiotica 77 (4): 441–459. refer to the first one as the semantic function, as it is related to what the text is about, and they call the other one the pragmatic function, as it is related to the reader. These two are coexistent, “with the semantic function being included in and justified by the pragmatic function” (Iarovici and Amel 1989Iarovici, Edith, and Rodica Amel, R. 1989 “The Strategy of the Headline.” Semiotica 77 (4): 441–459. , 442).

Research findings indicate that many readers spend more time scanning headlines (and looking for photos and dropped quotes) than reading articles, because this strategy maximizes their informational gain relative to invested cognitive effort (Dor 2003Dor, Daniel 2003 “On Newspaper Headlines as Relevance Optimizers.” Journal of Pragmatics 35: 695–721. ). This indicates that headlines often fail to induce the desire to read the article, not accomplishing one of the chief purposes they are intended to achieve. This might be a pressing problem in the case of internet headlines, whose effectiveness is measured in terms of click-through rates (see Kuiken et al. 2017Kuiken, Jeffrey, Anne Schuth, Martijn Spitters, and Maarten Marx 2017 “Effective Headlines of Newspaper Articles in a Digital Environment.” Digital Journalism 5 (10): 1300–1314. ), therefore the major concern of their authors is to ensure that readers are prompted to click on the provided link.

3.Clickbait headlines: Characteristic features

Research on headlines, which started many years ago as research into newspaper headlines (see Schneider 2000Schneider, Kristina 2000 “Emergence and Development of Headlines in English Newspapers.” In English Media Texts Past and Present: Language and Textual Structure, ed. by Friedrich Ungerer, 45–65. Amsterdam: John Benjamins. ), has recently expanded to cover headlines used on the internet, as this has become a growing market for news consumers. It is widely acknowledged that many more people worldwide are reading articles available online than paper copies (Kuiken et al. 2017Kuiken, Jeffrey, Anne Schuth, Martijn Spitters, and Maarten Marx 2017 “Effective Headlines of Newspaper Articles in a Digital Environment.” Digital Journalism 5 (10): 1300–1314. ; Bazaco et al. 2019Bazaco, Ángela, Marta Redondo, and Pilar Sánchez-García 2019 “Clickbait as a Strategy of Viral Journalism: Conceptualisation and Methods.” Revista Latina de Comunicación Social 74: 94–115. http://​www​.revistalatinacs​.org​/074paper​/1323​/06en​.html. ): in the busy life that everyone leads, it requires less time to pay brief visits to websites with news articles than to buy and read physical newspapers.

As indicated above, since opening the online article increases the click-through rates, which quantify the success of digital marketers and increase their revenue, enticing readers to access content and not just browse through headlines has become a primary concern of internet headline authors. As a result of the fierce competition for digital users’ attention, the phenomenon of clickbait headlines has become ubiquitous. This means that the pragmatic function appears to have acquired prominence these days and has become of greater importance than the semantic function of headlines, with the clickbait features of headlines recognized as the key element determining news readers’ choices on the internet (Kuiken et al. 2017Kuiken, Jeffrey, Anne Schuth, Martijn Spitters, and Maarten Marx 2017 “Effective Headlines of Newspaper Articles in a Digital Environment.” Digital Journalism 5 (10): 1300–1314. ).

There are various definitions of clickbaits (for a comprehensive survey, see Bazaco et al. 2019Bazaco, Ángela, Marta Redondo, and Pilar Sánchez-García 2019 “Clickbait as a Strategy of Viral Journalism: Conceptualisation and Methods.” Revista Latina de Comunicación Social 74: 94–115. http://​www​.revistalatinacs​.org​/074paper​/1323​/06en​.html. ). For the purposes of this paper, clickbait headlines will be defined as headlines on the web designed to attract attention in order to entice readers to follow the link provided and access the online text with a view to increasing the website traffic. Their ultimate goal, which has to do with triggering viral dissemination, is purely commercial. As Chen et al. (2015Chen, Yimin, Nadia Conroy, and Victoria Rubin 2015 “Misleading Online Content: Recognizing Clickbait as False News.” In Proceedings of the 2015 ACM on Workshop on Multimodal Deception Detection, 15–19. , 15) crudely put it, “the current state of online news is one that heavily incentivizes the speed and spectacle over restraint and verification in the pursuit of ad dollars”.

All this means that unlike good headlines, which are “not supposed to make the ordinary reader go on reading the story, but to ensure that the reader has indeed received the best ‘deal’ in reading the headline itself” (Dor 2003Dor, Daniel 2003 “On Newspaper Headlines as Relevance Optimizers.” Journal of Pragmatics 35: 695–721. , 718), clickbaits are devised to make the reader click on the story. Ifantidou’s (2009Ifantidou, Elly 2009 “Newspaper Headlines and Relevance: Ad Hoc Concepts in Ad Hoc Contexts.” Journal of Pragmatics 41: 699–720. , 700) argument that “[i]f headlines lack in informative value with respect to the article introduced, their function to attract attention may be more promising as a goal to fulfil” (emphasis added) is thus corroborated. Without exploring the issue of whether it is desirable for headlines to expose the most crucial information and, in effect, absolve the reader of consulting the text, which goes beyond the scope of the present paper, it needs to be underscored that the shift of their function in the digital environment to curiosity-raisers has become their most prominent characteristic (Kuiken et al. 2017Kuiken, Jeffrey, Anne Schuth, Martijn Spitters, and Maarten Marx 2017 “Effective Headlines of Newspaper Articles in a Digital Environment.” Digital Journalism 5 (10): 1300–1314. ).

The most conspicuous textual qualities of clickbaits include: sensationalism (reinforced by affective expressions and buzzwords), negativity and forward referencing (Blom and Hansen 2015Blom, Jonas, and Kenneth Hansen 2015 “Click Bait: Forward– Reference as Lure in Online News Headlines.” Journal of Pragmatics 76: 87–100. ; Kuiken et al. 2017Kuiken, Jeffrey, Anne Schuth, Martijn Spitters, and Maarten Marx 2017 “Effective Headlines of Newspaper Articles in a Digital Environment.” Digital Journalism 5 (10): 1300–1314. ). In their extensive quantitative study of clickbait headlines, Kuiken et al. (2017)Kuiken, Jeffrey, Anne Schuth, Martijn Spitters, and Maarten Marx 2017 “Effective Headlines of Newspaper Articles in a Digital Environment.” Digital Journalism 5 (10): 1300–1314. have found that in comparison with traditional headlines, clickbaits tend to include more signal words, sentimental expressions, citations and interrogatives. Typically marrying information with entertainment, clickbaits are classified as the so-called “infotainment” (Livingstone and Lunt 1994Livingstone, Sonia, and Peter Lunt 1994 Talk on Television. Audience Participation and Public Debate. London and New York: Routledge.). The information and type of content that headlines of this kind link to is quite diverse, as the examples below show:

Therefore, as Scott (2021)Scott, Kate 2021 “You Won’t Believe What’s in This Paper! Clickbait, Relevance and the Curiosity Gap.” Journal of Pragmatics 175: 53–66. rightly points out, it is not the nature of heralded content per se that is their distinctive characteristic, since spicy topics such as scandal, sex, the supernatural, etc. appear no longer enough to improve website traffic statistics. Their major distinguishing feature is that they induce readers to think that there is useful information to be learnt, thus creating what is referred to as an information gap (Loewenstein 1994Loewenstein, George 1994 “The Psychology of Curiosity: A Review and Reinterpretation.” Psychological Bulletin 116 (1): 75–98. ). Such expressions as “you need to know”, “you won’t believe” or “you’ve never heard of” explicitly communicate that the reader is deprived of some information that the text will provide. By indicating that “incredible tips” or “the secret ingredient”, etc. will be revealed, it is indicated that the knowledge to be gained is not commonly shared, that it is special and worth having and that by accessing it, the individual enters the privileged circle of those who are in the know.

There is unanimous agreement among researchers studying clickbait headlines (e.g., Blom and Hansen 2015Blom, Jonas, and Kenneth Hansen 2015 “Click Bait: Forward– Reference as Lure in Online News Headlines.” Journal of Pragmatics 76: 87–100. ; Potthast et al. 2016Potthast, Martin, Sebastian Köpsel, Benno Stein, and Matthias Hagen 2016 “Clickbait Detection.” In Advances in Information Retrieval, ed. by Nicola Ferro, Fabio Crestani, Marie-Francine ‎Moens, Josiane Mothe, Fabrizio Silvestri, Giorgio Maria Di Nunzio, Claudia Hauff, and Gianmaria Silvello, 810–817. Switzerland: Springer. ; Bazaco et al. 2019Bazaco, Ángela, Marta Redondo, and Pilar Sánchez-García 2019 “Clickbait as a Strategy of Viral Journalism: Conceptualisation and Methods.” Revista Latina de Comunicación Social 74: 94–115. http://​www​.revistalatinacs​.org​/074paper​/1323​/06en​.html. ; Scott 2021Scott, Kate 2021 “You Won’t Believe What’s in This Paper! Clickbait, Relevance and the Curiosity Gap.” Journal of Pragmatics 175: 53–66. ) that the decoy that they use is achieved through the information gap that they create. This means that their constitutive function, as hinted at above, is that of being curiosity-triggers. In particular, the mechanism employed has to do with promising readers that they will find out something that they need to or should know by clicking through, so they feed on the intrinsic Freudian ego-based need to know.

The tendency to close knowledge gaps and seek information is deeply rooted in human nature and is hypothesized to have developed as an adaptation to deal with uncertainty, so it has a firm evolutionary basis and biological underpinnings (Shin and Kim 2019Shin, Dajung, and Sung-il Kim 2019 “Homo Curious: Curious or Interested?” Educational Psychology Review 31: 853–874. ). The drive to seek information which helps to overcome knowledge deficiency is directly related to the general cognitive adaptation mechanism that aims at minimizing uncertainty and maximizing accuracy in representing the world around, crucial for survival (ibid.).

Information-seeking behaviours are thus built into human functioning, and clickbait headlines take advantage of that: the intrinsic purpose of clickbaits is to switch them on. Following Day’s (1982)Day, H. I. 1982 “Curiosity and the Interested Explorer.” Performance and Instruction 21 (4): 19–22. construal of curiosity, it can be contended that clickbaits let the audience into the “zone of curiosity”, at the same time providing an easy way out of this zone by pretending to make the potentially desirable information available to them at a click. In other words, the audience is led to believe that the dissonance between what they know and what, allegedly, they necessarily need to know can be resolved by getting acquainted with the content of the article provided. It is apparent then that curiosity resolution can be achieved at a low cost (see Sweeny et al. 2010Sweeny, Kate., Darya Melnyk, Wendi Miller, and James Shepperd 2010 “Information Avoidance: Who, What, When, and Why.” Review of General Psychology 14 (4): 340–353. ), so a favourable cost-benefit balance is supposedly guaranteed (Loewenstein 1994Loewenstein, George 1994 “The Psychology of Curiosity: A Review and Reinterpretation.” Psychological Bulletin 116 (1): 75–98. ; Shin and Kim 2019Shin, Dajung, and Sung-il Kim 2019 “Homo Curious: Curious or Interested?” Educational Psychology Review 31: 853–874. ). But the promise that by reading the content found at the landing site readers will satiate their curiosity is not delivered: what is signalled in the headline as “surprising” proves ordinary and familiar, what is proclaimed to be “incredible” turns out to be rather obvious and unspectacular, and what is denoted as “secret” appears common knowledge or trivial. So those lured into reading the text are bound for disappointment (Chen et al. 2015Chen, Yimin, Nadia Conroy, and Victoria Rubin 2015 “Misleading Online Content: Recognizing Clickbait as False News.” In Proceedings of the 2015 ACM on Workshop on Multimodal Deception Detection, 15–19. ; Scott 2021Scott, Kate 2021 “You Won’t Believe What’s in This Paper! Clickbait, Relevance and the Curiosity Gap.” Journal of Pragmatics 175: 53–66. ).

4.Deceptive clickbaits: A relevance-theoretic analysis

Inevitably then, the hopes of those who take the bait and click on the link are frustrated, as they find out that the content provided by the text that the headline links to was not worth their while. This will be upsetting for some readers and will dispel the illusions of others who have already had experience with this genre. But there are clickbait headlines that fool the reader in a more upsetting way, and it is these that the analyses in the present paper focus on. It is deceptive clickbaits and the manipulative mechanism they employ, which involves the careful staging of the interpretation process, that will be in focus here.

Deceptive clickbaits are manipulative headlines whose meaning is incongruent with the content of the text. In other words, consulting the article, readers soon find out that there is a discrepancy between what they thought the article would be about and the content they find on the landing site (Ecker et al. 2014Ecker, Ullrich, Stephan Lewandowsky, Ee Pin Chang, Rekha Pillai 2014 “The Effects of Subtle Misinformation in News Headlines.” Journal of Experimental Psychology Applied 20 (4): 323–335. ). It is not the case then that the disappointment stems from the fact that the promised upgrade of knowledge is not attained, as it happens with classic clickbaits referred to above. The problem is that there is a considerable mismatch between what the reader is led to think that the text is concerned with and its actual content. It is argued here that this incompatibility has its source in how the interpretation of the headline is manipulated, and in particular, in how the semantic underdeterminacy in the headline is resolved during the comprehension process. The point is that, as will demonstrated below, the way a given headline is created biases the reader towards a specific interpretation, arousing his or her curiosity. When working out what the headline communicates by assigning values to context-sensitive expressions, fixing reference, identifying the meaning of vague lexical items and phrases, etc., that is, performing all kinds of pragmatic tasks that verbal comprehension requires, the reader is deliberately led to the interpretation that arouses his or her intense curiosity, with a view to luring the individual into clicking on the link provided.

It needs to be highlighted at this point that it is not the case that the deceptive clickbait is a blatant lie (cf. Dynel 2018Dynel, Marta 2018 Irony, Deception and Humour. Seeking the Truth about Overt and Covert Untruthfulness. Boston/Berlin: De Gruyter. , 436). Simply, the headline, as it stands, is compatible with more than one interpretation, and the one that the reader is expected to access, critical to creating the information gap, is not consistent with the article’s content. Unlike lies, which are intended to produce a false belief in the target (Dynel 2018Dynel, Marta 2018 Irony, Deception and Humour. Seeking the Truth about Overt and Covert Untruthfulness. Boston/Berlin: De Gruyter. , 16), deceptive clickbaits merely misrepresent the content of the article by projecting a certain meaning which is later discovered to be unintended. This means that the reader is manipulated into thinking that the text is about a certain issue, while in fact it is about something else (on an in-depth study of the language of deception, see Galasiński 2000Galasiński, Dariusz 2000 The Language of Deception. A Discourse Analytical Study. Thousand Oaks: Sage. ). Ultimately, the reader’s expectations are frustrated; however, the goal of the clickbait has been achieved as long as the user decides to visit the site. The crucial question to answer is: how is the audience manipulated to interpret the headline in the intended way, so that the curiosity trigger that provokes information-seeking is released?

As will be shown below, a relevance-theoretic analysis provides a plausible and convincing explanation of what is involved. Before some examples are used to demonstrate how the underlying comprehension mechanisms work, certain fundamental assumptions and claims of relevance theory need to be presented. In very general terms, in this approach, verbal comprehension is postulated to consist in the formulation and evaluation of hypotheses about the contextually intended meaning on the basis of the evidence in the form of an utterance provided by the communicator for precisely this purpose (Sperber and Wilson 2002Sperber, Dan, and Deirdre Wilson 2002 “Pragmatics, Modularity and Mind–Reading.” Mind & Language 17: 3–23. , 7). The process of utterance comprehension, which, as is posited, starts with a decoded logical form as returned by the language parser, involves inferential adjustment and fleshing out of this semantic template, resulting in the full-fledged contextual meaning. Thus, verbal comprehension is by definition assumed to be inferential in this model (Sperber and Wilson 1995Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell.; Wilson and Sperber 2012Wilson, Deirdre, and Dan Sperber 2012 Meaning and Relevance. Cambridge: Cambridge University Press. ).

Since, as other human cognitive behaviours, comprehension is assumed to be relevance-driven (Sperber and Wilson 1995Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell.; Wilson and Sperber 2012Wilson, Deirdre, and Dan Sperber 2012 Meaning and Relevance. Cambridge: Cambridge University Press. ), in the process of utterance understanding the major goal of the hearer is to generate communicator-intended cognitive effects, which can roughly be understood as improvements in the information state of the individual (Jaszczolt 2012Jaszczolt, Katarzyna 2012 “Semantics/Pragmatics Boundary Disputes.” In Semantics: An International Handbook of Natural Language Meaning, ed. by Claudia Maienbom, Klaus von Heusinger, and Paul Portner, 2333–2360. Berlin: Mouton de Gruyter., 2346). A fast and frugal comprehension heuristic is postulated to be tacitly followed by comprehenders. In accordance with this heuristic, the interpreter is assumed to “[f]ollow a path of least effort in constructing an interpretation of the utterance (and in particular in resolving ambiguities and referential indeterminacies, in going beyond linguistic meaning, in supplying contextual assumptions, computing implicatures, etc.)” and to stop as soon as their expectations of relevance are satisfied (Sperber and Wilson 2005Sperber, Dan, and Deirdre Wilson 2005 “Pragmatics.” UCL Working Papers in Linguistics 17: 353–388., 360). It is argued that this kind of procedure secures an optimally relevant interpretation, “i.e. one providing the maximum of cognitive gains at a justifiable processing cost” (Piskorska 2020Piskorska, Agnieszka 2020 “Introduction: The Literal-Figurative Language Continuum and Optimally Relevant Interpretations.” In Relevance Theory, Figuration, and Continuity in Pragmatics, ed. by Agnieszka Piskorska, 1–22. Amsterdam: John Benjamins. , 2). In this model then, rapid and relevance-constrained inferencing is taken to be responsible for generating the intended meaning. Let us see how it works in the context of understanding clickbaits and first consider Example (6) below:11.The examples discussed here come from a private corpus collected by the present author and two other people, Agnieszka Piskorska and Alicja Feitzinger, who had been asked to send me links to web materials in which they detected an obvious mismatch between the headline meaning and the content of the linked text. Over a period of 4 months in 2020, 22 deceptive clickbaits were gathered. It must be added that the data were collected sporadically and randomly, as none of us is a regular news website browser. I would like to thank both friends for their help with collecting the clickbaits.

When reading this headline, the reader is likely to think that something terrible happened during a funeral, because of which the boy referred to died. This interpretation surfaces, since while recovering the explicitly communicated meaning, or explicature of (6), schematically represented as (6a),22.It is a convention in relevance-theoretic analyses to capitalize explicatures as formulas that show mentalese representations. The bracketed material stands for the intended referent, which at this stage of comprehension remains not fully resolved but is sufficient to secure the recovery of intended cognitive effects. The asterisks used indicate that occasion-specific senses, referred to in the relevance-theoretic framework as ad hoc concepts resulting from pragmatic modulation, are involved. For an in-depth discussion on how relevance theory approaches ad hoc concept construction, see, i.a., Carston (2002Carston, Robyn 2002 Thoughts and Utterances. The Pragmatics of Explicit Communication. Malden, MA: Blackwell. , 2010Carston, Robyn 2010 “Lexical Pragmatics, Ad Hoc Concepts and Metaphor: From a Relevance Theory Perspective.” Italian Journal of Linguistics 22 (1): 153–180.), Sperber and Wilson (2008)Sperber, Dan, and Deirdre Wilson 2008 “A Deflationary Account of Metaphors.” In The Cambridge Handbook of Metaphor and Thought, ed. by Raymond Gibbs, 84–105. Cambridge: Cambridge University Press. and Wilson and Sperber (2012)Wilson, Deirdre, and Dan Sperber 2012 Meaning and Relevance. Cambridge: Cambridge University Press. . the reader (following the path of least effort) fine-tunes the meaning of the phrase “during funeral” into “at a certain point in the course of the funeral”. This interpretation of the phrase is most accessible to the reader and yields a satisfying range of cognitive effects in context, hence it is optimally relevant, and it is automatically identified as the intended one.

As indicated above, the meaning communicated by the headline conveys that something extraordinary must have happened at the funeral in consequence of which the boy died, as generating the explicature gives rise to the recovery of additional implicit import. This is directly related to the satisfying range of cognitive effects that the interpretation process is taken to bring about. The idea that verbal comprehension has to do with obtaining satisfying cognitive effects needs to be glossed at this point. It is assumed in relevance theory that the cognitive effects that utterance interpretation yields usually include constructing hypotheses about both explicitly and implicitly communicated import: more often than not speakers convey meanings pertaining to these two layers of content, and the pragmatic module in the hearer’s mind is responsible for generating explicatures as well as implicatures (Wilson and Sperber 2012Wilson, Deirdre, and Dan Sperber 2012 Meaning and Relevance. Cambridge: Cambridge University Press. ). Implicatures calculated through inferencing alone, unlike explicatures that necessarily entail decoding and inferring (Sperber and Wilson 1995Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell.), are an important part of the interpretation process and have to do with fulfilling the specific expectations of relevance in a given communicative situation (for a more detailed discussion of this issue, see Sperber and Wilson 1995Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell.; Carston 2002Carston, Robyn 2002 Thoughts and Utterances. The Pragmatics of Explicit Communication. Malden, MA: Blackwell. ; Wilson and Sperber 2012Wilson, Deirdre, and Dan Sperber 2012 Meaning and Relevance. Cambridge: Cambridge University Press. ). These expectations in the context of the headline under inspection are likely to prompt implicatures, possibly conceptually similar to those in (6b) and (6c):

The implicatures like those above give rise to the questions about what exactly happened at the funeral (what was the accident at the funeral that killed the boy? how did he die?), so the information gap opens up and the reader is provoked into clicking on the link. In fact, the text reveals that the boy drowned while his parents were attending a funeral. A felicitous headline providing a true summary of the tragic event, for instance, the one in (6d) would not create the desired curiosity gap and might not improve clickability rates: fatal accidents concerning the young ones are more likely to cause sadness and distress than arouse widespread public interest.

In Example (7), the interpretation of the headline and its clickbaiting quality relies on the way the phrase “the Polish national team” is identified.

The relevance-theoretic model predicts that the reader interprets the title to refer to the current Polish football team’s coach, as schematically shown in (7a): as can be anticipated in a football-loving nation, this is the most accessible interpretation of the key phrase that is likely to occur to a Polish reader, since “the Polish national team” is commonly taken to stand precisely for the Polish national football team. Potentially rich in cognitive effects, this interpretation is derived by the comprehender following a path of least effort and surfaces as the intended one.

Obviously the news about the death of the Polish football team’s coach would be hot news, so again the kind of information gap that clickbaits are designed to fuel is created. The explicature in (7a) is likely to unleash a range of implicatures, possibly not unlike those in (7b)–(7d):

In fact, the text is about an ex-coach of Polish national junior volley-ball team, who passed away, so there is manifest incompatibility between the interpretation generated by the audience and the content of the news article.

The headline in (8) below signals that Tom Hanks has suffered extreme distress and sorrow in his life, suggesting that some catastrophic events from his biography are likely to be disclosed.

The word “tragic” is contextually interpreted to convey this. The expression “real-life” likewise suggests that the idealized picture of Tom Hanks will probably disappear after the reader learns about the facts that the story reports. As relevance theory predicts, the text of the headline will be optimally relevant yielding a satisfying range of cognitive effects on such an interpretation, which, in effect, will create an information gap and arouse curiosity in the audience, eventually making many of them click on the link.

However, the story is mainly about how Hanks faced bitter disappointments when his projects were not fully successful. It also underscores that his dad “worked long hours, often leaving the children to their own devices at home”. It is also reported that in 2013 Hanks was diagnosed with type 2 diabetes. The text exposes Hanks’ unfulfilled hopes and expectations, but all this can be hardly thought of as “tragic” in the sense that the reader expects: there are no truly horrible aspects of Hanks’ life to be found in the article.

As the examples above demonstrate, the relevance-theoretic model of utterance interpretation predicts which interpretation the reader will generate and how he or she is led to enter the “curiosity zone”. It explains how deceptive clickbait headlines bias the reader towards a specific meaning, which results in fostering the reader’s spirit of inquiry, tempting him or her into reading the text. However, when the news story is being read, the original interpretation of the clickbait is found incongruent with the content of the article. This is due to the fact that there is blatant incompatibility between what the reader is curious about and what the text is about, so his/her appetite for specific information is not satiated. In effect, the reader will feel disinformed by the headline. It may also occur to some readers that their first interpretation of what the headline meant was misguided, as they realize that the wording is compatible with another – less accessible but feasible and consistent with the news article – interpretation. This indicates that deceptive headlines are not strictly speaking lies (cf. Dynel 2018Dynel, Marta 2018 Irony, Deception and Humour. Seeking the Truth about Overt and Covert Untruthfulness. Boston/Berlin: De Gruyter. ), and the mechanism they employ is similar to that of joke punchlines.

5.Deceptive clickbaits and jokes

As Iarovici and Amel (1989)Iarovici, Edith, and Rodica Amel, R. 1989 “The Strategy of the Headline.” Semiotica 77 (4): 441–459. point out, the strategies of headlines resemble those of advertising. Both headlines and ads need to be attractive to the audience, create a desire in the recipient, and provoke and facilitate further action, which should be more or less immediate. As hinted at above, I would like to argue that there is affinity between deceptive headlines and jokes, which has to do with the way inferential interpretive processes are planned and exploited in the case of both of these text types.

As Yus (2016)Yus, Francisco 2016 Humour and Relevance. Amsterdam: John Benjamins. contends, in order to spark the humorous reaction in the audience, joke-tellers deliberately lead them along a certain interpretation path. Dynel (2018Dynel, Marta 2018 Irony, Deception and Humour. Seeking the Truth about Overt and Covert Untruthfulness. Boston/Berlin: De Gruyter. , ix) calls this a special “communicative design”, which in the case of canned jokes takes the form of garden-path interpretations (Dynel 2009Dynel, Marta 2009 Humorous Garden-Paths: A Pragmatic-Cognitive study. Newcastle: Cambridge Scholars Publishing.). It means that joke tellers predict and manipulate the way listeners/readers make sense of what they hear or read: the meaning the audience is led to recover while processing the joke’s set-up is invalidated when the punchline comes, and the resulting incongruity, which becomes finally resolved as appreciation of two contrasting meanings (ibid.), underlies intellectual satisfaction and humorous reaction in joke recipients.

Let us look at an example in order to illustrate how the manipulation whose goal is to create the humorous punchline effect operates.

The sense assigned to “waving her hair” in (9) will be that of “styling the hair so that it curls”, since this is the most salient and accessible interpretation: tacitly following the relevance comprehension heuristic, the reader is bound to understand it in this way. However, when the punchline comes, there is another sense of this expression that becomes manifest, namely – a rather absurd – meaning that has to do with using the hair as a flag and swinging it the way you wave a flag during public festivities. It is the incongruity that results from the two meanings which, as is postulated in Yus’ (2013Yus, Francisco 2013 “An Inference-Centered Analysis of Jokes: The Intersecting Circles Model of Humorous Communication.” In Irony and Humor: From Pragmatics to Discourse, ed. by Leonor Ruiz Gurillo, and M. Belén Alvarado-Ortega, 59–82. Amsterdam: John Benjamins. , 2016Yus, Francisco 2016 Humour and Relevance. Amsterdam: John Benjamins. ) Intersecting Circles Model of humorous communication, will trigger a humorous reaction.

The same mechanism is in operation in (10) and (11) below. In (10) the phrase “go to bed” takes on a totally different meaning when the punchline comes from the one originally assigned to it in the opening question, and in (11) the unexpected meaning of the priest’s question (blatantly incongruous with the most salient one) becomes manifest when the little girl answers the question the way she does.

As these examples demonstrate, there are clear parallels between how jokes achieve the desired effect and how deceptive clickbaits do. Both jokes and deceptive clickbaits are special kinds of texts that exploit interpretation bias: recipients of both jokes and clickbaits are supposed to follow a preconceived inferential path, with the inferential steps that the audience will take, predicted and consciously manipulated by their authors. The text of the joke is cleverly crafted to create an incongruity configuration that is the cornerstone of humour (Yus 2016Yus, Francisco 2016 Humour and Relevance. Amsterdam: John Benjamins. ). Analogously, the deceptive clickbait is designed in such a way that readers are biased towards the interpretation that opens the information gap, which by arousing their curiosity, makes them click on the linked text. While the audience of a joke is guided to the initial interpretation by the set-up of the joke, which creates conditions for the punchline effect (with a view to making the audience laugh), the deceptive clickbait manipulates the audience to recover the curiosity-arousing interpretation (with a view to alluring them to click), by making it the most salient and accessible one to the reader; this intended interpretation proves incompatible with the article the clickbait links to. The predictability of how a given joke and a deceptive clickbait will be processed has to do with resolving the underdeterminacies involved for both genres to achieve the strategic goal. Joke recipients become aware of two competing interpretations when the punchline comes, while those who have fallen prey to a deceptive clickbait discover incongruity between their interpretation of the clickbait headline and the text in the course of reading the accessed article.

There is a major difference between jokes and deceptive clickbaits though: while the former aim to amuse and bring forth satisfaction, the latter are meant to increase digital traffic (and ultimately bring economic rewards to platform owners) and the reaction of the fooled reader is usually disappointment and the feeling of being tricked. The intellectual pleasure that humorous texts culminate in contrasts with displeasure that deceptive clickbaits may create in the reader. This means that whereas jokes lead to a humorous climax, which can be explained in terms of an inferential overload effect (Jodłowiec 2015Jodłowiec, Maria 2015 The Challenges of Explicit and Implicit Communication. Frankfurt: Peter Lang. ; Piskorska and Jodłowiec 2018Piskorska, Agnieszka, and Maria Jodłowiec 2018 “Weak Communication, Joke Targets and the Punch-Line Effect: A Relevance-Theoretic Account.” Studies in Polish Linguistics 13 (1): 25–44.), deceptive clickbaits bring about an anti-climax, when the hopes of satisfying the aroused curiosity are dashed, and readers find out that they rose to the bait.

The punchline effect and the clickbait effect, while relying on the same interpretation mechanisms, are strikingly different. The punchline effect, as argued at greater length elsewhere (Piskorska and Jodłowiec 2018Piskorska, Agnieszka, and Maria Jodłowiec 2018 “Weak Communication, Joke Targets and the Punch-Line Effect: A Relevance-Theoretic Account.” Studies in Polish Linguistics 13 (1): 25–44.), has to do with an inferential overload effect: a wide range of weakly communicated assumptions suddenly becoming accessible in the recipient’s mind, leading up to a sudden reorganization in the background assumptions available. This can be easily traced as the cognitive effect that comes about when the punchline is processed in the examples presented above. For instance, in (11) there is a host of assumptions that become manifest to the joke recipient when the punchline comes, which concern naiveté and straightforwardness of little girls, promiscuity of male villagers, inquisitiveness of priests, and so on. All this takes place at the subpersonal level (which explains why we usually find it very difficult to explain what makes the punchline funny) and there will of course be differences in what becomes manifest to different individuals exposed to the joke.

No inferential overload occurs when the reader discovers that the text that he or she has just accessed is incompatible with their interpretation of the headline, which has proved to be deceptive. The implicatures generated by the original reading of the headline simply get cancelled, and there is no rich impact of the weakly communicated assumptions that would suddenly become manifest to the reader. The individuals tricked into clicking on the link simply realize they will not find what they have expected, which will result in disappointment. The point is that incongruity that functions as the inferential overload trigger in the case of jokes, becomes the cognitive impact killer in the case of clickbaits.

It may be expected that after a number of rather frustrating experiences with deceptive clickbaits, some readers will become cautious and their vigilance mechanisms, in particular those that Padilla Cruz (2015Padilla Cruz, Manuel 2015 “On the Role of Vigilance in the Interpretation of Puns.” Humor – International Journal of Humor Research 28 (3): 469–490., 2016Padilla Cruz, Manuel 2016 “Vigilance Mechanisms in Interpretation: Hermeneutical Vigilance.” Studia Linguistica Universitatis Iagellonicae Cracoviensis 133 (1): 21–29.) refers to as hermeneutical vigilance may be activated. As Padilla Cruz (2016)Padilla Cruz, Manuel 2016 “Vigilance Mechanisms in Interpretation: Hermeneutical Vigilance.” Studia Linguistica Universitatis Iagellonicae Cracoviensis 133 (1): 21–29. suggests, hermeneutical vigilance is responsible for checking plausibility and acceptability of interpretive hypotheses and makes hearers sensitive to misinterpretation. In the context of deceptive clickbaits there is a chance that readers frequently lured into deceptive traps by internet headlines will become attentive in this respect, and may develop vigilance against baits of this kind and (at least sometimes) refrain from clicking. However, it should be taken into account that it may not be very easy to distinguish good headlines from teaser ones. The authors of the latter are more and more careful to produce titles that would not signal sensationalization, so they are trying hard to camouflage deception. Exploring the issue of whether internet readers develop strategies to guard against rising to the bait seems an interesting line of further investigation.

6.Conclusion

There have been numerous studies on the language of newspaper headlines, but clickbaits have not received much attention from linguists so far (Scott 2021Scott, Kate 2021 “You Won’t Believe What’s in This Paper! Clickbait, Relevance and the Curiosity Gap.” Journal of Pragmatics 175: 53–66. ), with most research on clickbaiting conducted within the Artificial Intelligence paradigm. By focusing on the pragmatic analysis of deceptive clickbaits, the present paper attempted to fill at least a part of this gap in linguistic studies.

Out of several functions originally associated with headlines, one has become particularly conspicuous and significant in the digital age, namely, to attract attention. This is precisely the function that clickbait headlines are intended to serve: their major goal is to ensure high rates of clickability, and in order to achieve this, such headiness promise much more than they actually deliver. Deceptive clickbait headlines misguide readers as there is incongruity between the interpretation of the headline that the recipient is intended to recover and the content that the clickbait links to. Strictly speaking though, the headline does not provide false information: the way it is formulated makes it open to competing interpretations and the responsibility for recovering a certain meaning, though cleverly “engineered” by the website editor, rests with the reader.

My aim was to show how the relevance-theoretic model of utterance interpretation can be used to explain the mechanism underlying deceptive clickbaits: it provides useful insights into the nature of the bait and elucidates how the reader is enticed to click. In particular, predicting how the interpretation of the clickbait will proceed, the relevance comprehension heuristic explains the way the information gap is created. As has been argued above, there are some interesting parallels between the manipulative strategy employed in deceptive clickbaits and in jokes, though the outcomes of processing the two types of texts are very different: the intellectual satisfaction associated with getting the punchline sharply contrasts with the disappointment of the clickbait victim. The different outcomes of jokes and clickbaits are likely to have significant social consequences, suggesting a path of future research on clickbait headlines.