Oral self-translation of stand-up comedy and its (mental) text: a theoretical model

1 Introduction

Stand-up comedians are increasingly choosing to perform in more than one language, both in response to situations of migration and to adapt to an ever more globalized comedy industry. This phenomenon raises the questions of whether stand-up comedians translate their own material across languages, of what challenges they face in doing so and of how they address these challenges. I myself have been performing comedy in both English and Italian for the past four years, and have used this experience as the basis for an investigation on these issues. To this introspective starting point, I then added an observational aspect, namely interviews with other bilingual stand-up comedians. In particular, I conducted nine in-depth semi-structured interviews: four with fellow Italian comedians who regularly perform on the London circuit in both English and Italian (Giada Garofalo, Romina Puma, Luca Cupani, Federica Bonomi), two with Italian comedians based in Italy who performed in English in the UK (Francesco De Carlo and Antonello Taurino) and three with non-Italian comedians with a wide experience of performing in two languages (Katsura Sunshine, Magnus Betnér and Katerina Vrana). These interviews were taken in person, recorded and, in some cases, translated into English. The transcription/translation was then sent to them for their approval. A consent form was also sent, which included the consent to be identified by name in publications deriving from this research. While full transcriptions and full thematic analyses will be made available in forthcoming publications, this article only contains few excerpts (from participants Giada Garofalo and Romina Puma) that were considered pertinent to the problem of the orality and/or textuality of stand-up comedy.

In particular, this article offers an outline of the theoretical model I developed to make sense of the phenomenon investigated. The relationship between the empirical evidence collected from the interviews and the elaboration of this model followed the path of an iterative bidirectional process, as suggested, for instance, by Hammersley (1989) as part of his inductive analysis methodology: a first elaboration of this model informed the choice of questions, while later the answers helped me to revise the model. In particular, the starting point was the perception, first derived from self-observation and later corroborated by the interviews, that while stand-up comedy is essentially an oral form of communication, at the same time it is a form in which some kind of text should be recognized in order to justify what is constant in a series of performance of the same “thing” for instance of the same Edinburgh Fringe Festival show. This has important implications for translation, which needs to be a translation of something. If translation is taking place in stand-up comedy, then, an applicable theory will require a notion of text capable of accounting for what is repeatable and constant in a series of oral performances, while also accounting for what is not, i. e. the text’s openness to variation and improvisation. Theoretical accounts of stand-up comedy currently available in the literature, on the other hand, either negate that a text-based approach is relevant to its study (e. g. Rutter 1997) or attribute to the text of stand-up comedy a rigidity that is actually specific to other comic genres, such as canned jokes (e. g. Attardo 2001). Once this new theoretical model of stand-up comedy text is outlined, a model of the phases involved in its oral self-translation can also be attempted. While the first model aims to fill a gap in humor studies, the second aims to fill a gap in translation studies. Existing studies in the latter field, for instance Akai (1997) and Wilson (2011), focus exclusively on what we might call oral-to-written self-translation. Bilingual stand-up comedy, on the other hand, offers the opportunity to study a form of self-translation that is oral in both its source and its target.

2 Stand-up comedy: an oral form of expression and its oral self-translation

The starting point of the present investigation is the recognition that stand-up comedy represents an oral form of expression: it is, indeed, not possible to assume the presence of a written text as the point of departure for a series of performances. This seems to be corroborated by the data collected from the point of view of translation. For instance, Giada Garofalo declares, “I start from something, what I don’t do is starting again from the script, I don’t translate the ‘original’ script. Sometime I improvise the translation directly on the stage.”

Even when a written script is present, it seems to play a very limited function, as it shown in this passage, in which Giada is asked by me (“GP”) to comment on the script of her show for the Edinburgh Fringe Festival 2013:

GP – It is written as if to be performed word by word. Did you follow it word by word in the performance too?

GG – Not at all, for instance there are entire chunks of it that I actually never performed. Other parts were changed, without me updating the written script accordingly.

GP – What if you were asked to translate the show into Italian? Would you also write it word by word in the Italian version?

GG – No, maybe two years ago, but not now.

GP – What would the starting point be, then? The show how you developed it in the meantime?

GG – Yes, of course, there would be no point in “going backwards.”

Giada Garofalo’s starting point in this (although hypothetical) translation would, then, be something that in that form was never fixed on paper. Moreover, what might at first sight appear as pre-performance scripts in fact represent, in many cases, post-event transcripts of a specific performance. An example in this sense is offered by comedian Lee (2010), who offers to the reader a collection of entire show transcripts, commented in great depth by the comedian himself by means of very detailed footnotes. These footnotes make clear that many passages are, for instance, just “a transcript of what happened in one occasion” (2010: 182) and that “[a]ll this stuff was different every night” (2010: 265). On the contrary, written comedians’ notes are indeed antecedent to the performances, but not all comedians make use of them. For instance, comedian Jay Leno declares: “I have nothing written down […] I just keep it in my head. The good jokes I remember, the bad ones I forget” (Ajaye 2002: 122–123). Even when present, written texts are simply used as mnemonic tools. This view is consistent with the latest developments in the debate about the nature of orality itself.

While earlier studies, as exemplified by Ong (1982), saw oral forms of expressions as radically different in nature from written forms, Carr (2005) proposed a new paradigm, which he called “oral-written interface” (2005: 4). This approach is based on interpreting the writings of the ancient Greeks, Mesopotamians, Egyptians, and of the people of ancient Israel, as tools to enable the oral transmission of the texts, particularly by assisting memorization: writing was mainly seen as “writing on the tablet of the heart” (2005: 127). The same can be said about the specific form of “writing” behind stand-up comedy performances.

If stand-up comedy is an oral form of communication, making sense of the phenomenon of stand-up comedians translating their own material requires a specific notion of translation: I will call it oral self-translation as, in translation studies, the term self-translation is used to label the type of translation in which the author and the translator happen to be the same person. As mentioned in the introduction, the literature does not recognize what we might call oral-to-oral self-translation, i. e. of a form of self-translation where an instance of oral discourse in the source language is translated into an equally oral instance of discourse in the target language.

This oral-to-oral linguistic transfer appears to be what bilingual stand-up comedians do with their own material. The evidence collected in this study from the interviews seems, indeed, to suggest that the need to “write things down” as part of the translation process is only felt when the linguistic competence is weak and is felt by the comedians themselves as a limitation. A case in point is offered by Romina Puma, who reported the following:

[the passage to English] was three years after moving to London and it was terrifying! In Italian I could write down some “bullet points” only and improvise from there, but I didn’t trust myself to do the same in English, instead I had to write down and memorize every single word, as we used to do at school with those poems by Giacomo Leopardi! That meant that, on stage, I was terrified, since, for instance, I knew I wouldn’t be able to respond to hecklers.

This was not a problem, on the other hand, for previously quoted comedian Giada Garofalo when she prepared her show for the Edinburgh Fringe Festival 2013, at which time she had been living in England for eleven years. When linguistic competence allows, then, the text of stand-up comedy is developed as an oral text and its translation appears to be a form of oral-to-oral translation.

3 The concept of mental text and its declarative/procedural nature

Once established that there might be such a thing as oral-to-oral self-translation, the question of what can be identified as its source text and target text arises. In the case of oral self-translation of stand-up comedy, this points to the more general question of what constitutes the text of stand-up comedy itself. Rutter (1997) underlines the conversational nature of stand-up comedy and suggests that “[t]raditional text-based analyses cannot record, and therefore facilitate, the exploration of the interactive nature of humor” (1997: 291). Rutter is arguably right in underlining the importance of interaction in stand-up comedy. The question, however, is what the comedians bring to this interaction, apart from their personality, history and attitudes. Indeed, it is not as a blank slate of conversational intentions that the comedians approach the stage. On the contrary, they approach it with the intention of having a specific type of conversation, a conversation with a specific content. If it is true that “it is how you say it”, to quote a well-known adage about jokes, there must also be an “it” that is said.

Rutter does not deny all this, but considers the content of what is said less relevant than other aspects of the stand-up comedy “conversation”. An example he cites as evidence to this point is the episode of a Big Issue reviewer joining the general laughter for a joke during a performance by Johnny Vegas, and reporting this experience in his review along with a transcription of the joke, despite completely misunderstanding the content of the joke itself (1997: 196). While interestingly revealing of the contagious and social nature of laughter (and, maybe, of the level of competence of some comedy reviewers), this example clearly represents an extreme case and can only be viewed as an exception to what occurs in the comedy room. It will be difficult, in fact, to deny that behind the laughter from most people in the room there needs to be at least some common understanding of what is said; otherwise, the very conversational nature of stand-up comedy would be lost, since it would be a conversation apropos of nothing.

If the search for the text of stand-up comedy is relevant and pertinent, the quote from Jay Leno reported above can now be read as providing a positive suggestion on where to locate it. While negating the existence of something “on paper”, Jay Leno assures that there is something “in the head”. This remark has an aura of familiarity with what was suggested by ethnographer Barber, according to whom in oral performances “[s]omething identifiable is understood to have pre-existed the moment of utterance […] [that] can be held to exist is in people’s minds or memories” (2005: 325).

Indeed, the notion that a text can be located “in the mind” is not completely new. In his study about the ontological status of texts, philosopher Gracia (1996) introduces the notion of mental text as a separate entity from the physical text. According to the theory presented in his study, however, a mental text is still dependent on the presence of a written text: it is the mental image of a written text (1996: 23), which either follows it, for instance in the reader’s mind during the act of reading, or precedes it, for instance in the writer’s mind during the act of writing (1996, 24: 25). The mental text, then, does not necessarily appear a posteriori in the mental reproduction of the written text, but can also appear a priori in its production. To recall Chomsky (1965), the notion of mental text, consequently, seems to acquire the aspect of a form of competence which underlies and makes possible the performance of the concrete act of writing the text. It would not require, at this point, a great effort of theoretical imagination to extend this notion to the idea of a mental text that underlies and makes possible a series of oral performances.

This very extension, although apparently in total autonomy from Gracia (1996), has been proposed by ethnographer Honko (1996) in order to describe the “something” behind the performances of singers of epics along the Silk Roads:

To be able to understand the production of text in actual performance, it seems necessary to postulate a kind of “pre-narrative,” a pre-textual frame, that is, an organized collection of relevant conscious and unconscious material present in the singer’s mind. This material consists of (1) textual elements and (2) generic rules for reproduction; we may call it a “mental text.”

(Honko 1996: 1)

This notion of a mental text might raise methodological concerns, since it seems to locate the text in a realm that is not open to direct investigation, as observed by Claus (2000). Fortunately, researchers on self-translation of stand-up comedy enjoy an obvious advantage over people who study singers of oral traditions along the Silk Roads: a much easier access to the performers themselves. Interviews, introspection and, more generally speaking, a phenomenological approach to the “experience” of translating become, consequently, central to this type of investigation.

The mental text, then, first of all seems to be something that is “present in the singer’s mind” (Honko 1996: 1), or more generally the performer’s mind, across time, since typically it needs to enable not only a single performance, but a series of performances. If, for instance, potential audience members read a review of the fifth performance of the festival run of a comedy show and decide to go and see its seventh performance, this decision is only justifiable under the assumption that there is something that is permanent across all performances. As Barber puts it: “it is clear that what happens in most oral performances is not pure instantaneity, pure evanescence, pure emergence and disappearance into the vanishing moment” (Barber 2005: 325). The mental text, then, is something that is stored. Similarly, Attardo mentions the need for a theory of humorous texts to recognize “the presence of a ‘storage area’ for the information that is being assumed, shared and developed by the text” (1996: 88). The abstractness of the term “storage area” is probably due to the fact that the paper was originally developed in the course of a workshop about computational humor, but in the all too human context of this article the term “memory” is sufficiently comprehensive. An investigation on memory in oral tradition has been attempted by Rubin (1995), although with a specific interest for mnemonic techniques. In more general terms, talking about memory allows to take advantage of the considerable literature dedicated to this subject in the fields of cognitive psychology and neurology, for instance, with regard to the classification of memory systems (for an overview: Squire 2004) and the different types of content associated with them. Indeed, the second aspect of the mental text described in the passage quoted from Honko is, so to speak, its double nature: a mental text is constituted both of “textual elements” (Honko 1996: 1) and by “rules of production” (Honko 1996: 1). If we consider, then, that the mental text is stored in memory, the question that arises is whether these different types of content are stored in different types of memory. Answering this question would allow a better understanding of what the contents of a mental text might entail.

Many authors in the field of cognitive science distinguish between declarative and procedural memory (for a brief literature review see: Ullman 2004). First, declarative memory “refers to the capacity for conscious recollection about facts and events” (Squire 2004) and “can be divided into semantic memory (facts about the world) and episodic memory (the capacity to re-experience an event in the context in which it originally occurred)” (Squire 2004: 174). On the other hand, procedural memory is the type of memory that “subserves the learning of new, and the control of established, sensori-motor and cognitive ‘habits’, ‘skills’, and other procedures, such as riding a bicycle and skilled game playing” (Ullman 2004: 237). Its learning and recollection is mostly unconscious and, for this reason, is also referred to as “implicit memory” (Ullman 2004: 237).

The latter point shows why the mental text of stand-up comedy cannot be considered as exclusively procedural. If it were so, performing a comedy set in a different language would just be akin to riding a different bike, to use a well-known metaphor of procedural skills. If this was the case, however, there would be no need for any conscious translation and the (unconscious) translation would just happen “on the stage”, not prior to the performance, since chronological precedence would require conscious recollection. The phenomenological evidence collected both from the interviews and from my own introspection, however, seems to be in strong contradiction with this hypothesis: comedians do seem to ask themselves questions of translation, which are both conscious and chronologically antecedent to the performance itself. For example, participant Giada Garofalo reported in her interview the following:

when the punch line of a joke happens to be culture-specific or language-specific […] I need to spend more time on finding an equivalent in the other language, in advance to the first performance in it.

On the other hand, if the mental text of stand-up comedy was stored exclusively in declarative (semantic) memory it would not be any different from the script of a play. What would not be accounted for, in that case, would be the ability of the comedian to improvise asides, to counter-react to the audiences’ reactions, to respond to hecklers, to recognize what comedians call “the elephant in the room” (latecomers, noises, uncomfortable environment, news that are likely to be in everybody’s mind). There is, it appears, the need to recognize that the comedian approaches the stage not only with (declarative) contents in mind, but also armed with (procedural) skills.

The failure to make sense of the mental text of stand-up comedy in either exclusively procedural terms or in exclusively declarative terms points to the recognition of its double procedural/declarative nature. This recalls Ullman’s (2004) hybrid declarative/procedural model of language, in which linguistic competence itself is described as constituted by both declarative (for instance, lexicon) and procedural (for instance, grammar) elements. Moreover, this double procedural/declarative nature of the mental text, and the distinction between these two aspects, might also shed some light onto the relationship between the content of oral discourse and the mode of its production in performance, in other words, the “how” in the customary observation that “it’s how you tell it”. The performative aspects of a mental text can be seen as one of the main constituents of its procedural part. What seems to be the case, then, is that a “stack” of procedural skills can be recognized as being deployed in performance: from the most generic linguistic competence described by Ullman (2004), through the “generic rules of production” described by Honko (1996), to the specific skills associated to a particular content, such as how to make a specific punch line more effective by underlining it with a facial expression.

4 Zooming in on the declarative content of the mental text of stand-up comedy: the punch lines

While the content of the procedural portion of a mental text will inevitably need to be left to some extent indeterminate (it is, after all, the part of the model called to justify what is less predictable in the performance of the text), it is now time to focus on the content of the declarative part. The claim is that it consists, mainly, of jokes and/or in punch lines. This claim, however, poses a series of questions.

First of all, phenomenological evidence, both based on self-observation of my experience in delivering stand-up comedy in two languages and on the interviews collected for this research, seems to suggest that comedians dedicate great attention to the wording, in both the source language and the target language, of those parts of discourse from which they expect a humorous response. There seems to be much confusion, however, both in the literature and in common use, on what terms best describe these essential parts of humorous discourse, particularly with regard to the distinction between jokes, canned jokes and punch lines (sometimes called “punchlines”). A brief discussion about this distinction is then needed.

Attardo (2001), for example, lists among the distinguishing features between canned and conversational jokes both conventional elements of their narrative structure, such as “an announcement of the humorous nature of the forthcoming turn” (2001: 61), and elements of a more cognitive nature, such as that these jokes are rehearsed (2001: 62). Consequently, Attardo then declares that “a stand-up routine is a highly rehearsed, planned text, which consists in a (sometimes large) part of canned jokes” (2001: 62). Applying Attardo’s own definition, however, would lead to the expectation of finding announcements of the humorous intention repeated all over the stand-up comedy routine. Although Attardo himself qualifies the frequency of the production of these announcements of intention at the beginning of a canned joke as happening “often” (2001: 62), not for every canned joke, there is still a contradiction with the evidence suggesting that these announcements are actually very rare in stand-up comedy. This observation is mainly based on personal experience, both as a performer and an audience member of stand-up comedy, but some independent corroboration can be found in the analysis of show transcripts, for instance those collected in Lee (2010). It is, in fact, the stand-up comedy situation itself that frames the discourse as containing humorous remarks. If these narrative strategies and structures are used as the defining characteristic of canned jokes, then, stand-up comedy sets do not seem to be constituted of them.

Thus, there seems to be a gap in the terminology of humor studies, i. e. the lack of a term to indicate a joke that, on one hand, does not occur spontaneously in conversation (so it represents a canned joke in the cognitive sense) and, on the other hand, is not an instance of a conventional text type (it is not a canned joke in the narrative sense). Stand-up comedy jokes would appear to be instances of this type. What both stand-up comedy jokes and canned jokes (in their stricter, narrative sense) have in common, however, is that they end in punch lines (at least under the definition of punch line that will soon be suggested). The difference, then, can be identified in the different textual strategies put in place to prepare for the delivery of the punch line. Focusing on punch lines, then, offers the advantage of avoiding the conceptual and terminological quagmire around the notion of joke. Moreover, it appears consistent with the main concern of the performers, particularly with reference to the task of memorization. For instance, Davies (2008) notes that “[m]ost good joke-tellers do not memorize jokes. They simply remember the punch-line [sic]” (2008: 160).

An investigation on the nature of punch lines, then, becomes of the greatest importance for the understanding of the declarative portion of the mental text of stand-up comedy and, hence, of the object of the conscious oral self-translation activity. Oring (1989) notes that “[t]he notion of the punch line remains largely intuitive” (1989: 351), suggesting a definition from within the framework of an incongruity theory of humor, particularly the script theory proposed by Raskin (1985): “The punch line is a device that triggers the perception of an appropriate incongruity” (1989: 351). Later in the same study, Oring adds that “[a] punch line must stand in the final position” (1989: 353), to the point that if this rule is violated “the narrative will tend to be conceptualized less a joke and more as a form as of humorous tale” (Ibid). The problem seems, again, to consist in the semantic confusion about the meaning of the word “joke”: Oring seems to identify jokes with canned jokes in the narrative sense, using instead humorous tales for instances of (still non-spontaneous, we can assume) humor discourse that do not fall within this strict definition. Going back to the discussion about the position of punch lines, Attardo (1996) accepts the restriction of punch lines to final position of a canned joke and, consequently, feels the need to introduce the concept of jab lines (1996: 91), specifying that “[j]ab lines differ from punch lines in that they may occur in any other position in the text” (Attardo 2001: 82). The need to introduce this distinction, however, appears to be the result of the enforced superimposition on all humorous texts of the narrative conventions associated to canned jokes in the stricter, narrative sense (as opposed to the cognitive sense, of jokes that are rehearsed). It is only, in fact, as an effect of considering stand-up comedy as composed by a sequence of canned jokes in the narrative sense that the distinction between punch lines at the end and punch lines in any other position, i. e. jab lines, can make some sense. If, instead, we abandon the idea of stand-up comedy as composed by a sequence of canned jokes in the narrative sense (although they can still be considered such in the cognitive sense, i. e. they are usually not spontaneous), this distinction disappears, since there is now no narratively recognizable smaller unity of text (within the enclosing comedy set) to be at the end of.

The objection that not all stand-up comedians make use of punch lines, which might arise from the previous comments, can also be seen as rooted in this strict definition of the notion of punch line, which in turn is derived by the superimposition on stand-up comedy of the rules of canned jokes. It seems that, regardless of whether the comedian’s discourse can or cannot be reconstructed as a sequence of short stories with a clear-cut ending, there will always be sentences within it to which a humorous intention is more obviously attached. Punch lines, then, will be those sentences to which comedians attach the intention of producing a humorous effect. This definition allows for an application in which the punch line seems to be absent, such as (in the world, this time, of narratively defined canned jokes) the so-called “shaggy dog stories”, in which the comic effect is derived by the fact that the expectation produced by a very long set up is reversed by the delivery of some trivial sentence. It is clear that the narrator intends this sentence to produce the humorous effect.

From the point of view of an external observer, punch lines will probably be recognized as such by observing the audience response. This approach, however, entails difficulties in recognizing a failed punch line – which is a problem that in turn has consequences for translation, as will soon be discussed. Fortunately, this study focuses on the phenomenon of self-translation, in which introspective access to somebody’s own intentions can be reasonably assumed.

5 Focusing on the bilingual perspective: a model for the translation of the mental text of stand-up comedy

In Section 4, the discussion on the content of mental text has led to identifying in the punch lines the most important constituents that are memorized declaratively. In this section, it is suggested that the (conscious and precedent to the performance) translation of punch lines represents the main task involved in the oral self-translation of stand-up comedy. More specifically, it is one of the defining characteristics of declarative memory that “at least part of this knowledge can be consciously (‘explicitly’) recollected” (Ullman 2004: 235). The empirical evidence collected in this study seems to confirm that at least a portion of the translation effort is conscious and performed on consciously recalled material. If this is the case, it can only have as its object the part of memory that is available for conscious recollection, i. e. declarative memory. Having identified in punch lines the minimal, and functionally most important, constituents of the declarative content of the mental text of stand-up comedy, the conclusion is that punch lines will also constitute the main focus of the conscious and pre-performance translation activities. On the other hand, content-specific procedural skills, such as the production of the most effective facial expression to underline a specific punch line, will be simply re-deployed for the target audience and in the target situation, which will represent the new “bike” that the performers will “ride”. This does not mean that they will not produce, in the example, a different facial expression. Considerations of the different value associated with expressivity in the source and target culture might convince the performers that, for instance, a more explicit expression is expected by an Italian audience than by a British audience. This sort of “translation”, however, will happen directly on the stage and unconsciously.

The declarative part of the mental text of stand-up comedy, instead, will be translated consciously. In order to gain a better understanding of how this might happen, a theoretical model of the different phases involved in this process can be suggested. A useful concept for this is that of mental model developed by Johnson-Laird (1983). According to Johnson-Laird, discourse comprehension, and understanding in general, consists in deriving from the discourse situation a mental model of some state of affairs. When discussing how this type of semantics can be applied to propositional attitudes, such as beliefs, Johnson-Laird suggests that a “propositional attitude is a relation between an individual and that individual’s mental model of the relevant state of affairs” (1983: 433). To understand “X believes that snow is white” we will, first, construct a mental model representing the state of affairs “Snow is white”, then we will construct a mental representation of the agent “X” and, finally, we will construct a mental representation of the relation of believing directed towards that mental model by that agent. Generally speaking, we will need to embed the mental model of the state of affairs expressed by the proposition P into a mental model of the relation towards it, i. e. of X believes that P.

A similar embedding can be hypothesized in those situations in which the main goal is not the understanding of “what is said”, but of the discourse situation itself. According to Setton (1999), for instance, this is what happens in simultaneous interpreting. Applying to this situation the theoretical idiom of mental models, Setton suggests that the interpreter constructs a mental representation not only of the speaker(s), but also of the audience, of their utterances and of the effect that the speakers are understood to want to produce on their audience by means of these same utterances. If the speaker is saying “Snow is white”, the simultaneous interpreter will produce a mental model that includes not only the (embedded) mental model representing snow being white, but also the (embedding) mental model of the speaker’s intentions in producing such utterance, which also includes a representation of the audience. Along similar lines, de Jongste (2016) offers an explanation of the recognition of the speaker’s humorous intent based on the perception of a shift between the mental model publicly expressed by the speaker and the audience’s reconstruction of what de Jongste calls his/her “private mental model” (2016: 108). For instance, in de Jongste’s example (2016: 111), when comedian Eddie Izzard pretends (publically) to believe that the humming sounds of Star Wars derive from Gregorian chant, which the audience know it is unlikely to be a (privately) held belief.

While constructing the mental model of comedic discourse, however, comedians are not only concerned with the perception of their humorous intent, but also with the conditions of its success. In order to understand better what might be involved in the construction of such a model, it might then be useful to point out that, according to Johnson-Laird, “the mental model is constructed on the basis of truth conditions of the propositions expressed by the sentences in the discourse” (Johnson-Laird 1983: 407). This application of the concept of truth conditions, however, seems to be problematic in the case of those propositions that Austin (1962) and Searle (1969) called speech acts, for instance commands. Singh (1993) suggests that this problem would be resolved by identifying the semantics of speech acts not with its truth conditions, but with its satisfaction conditions, defined as “the conditions under which we would affirm that the given speech act had been satisfied, e. g. a command is satisfied if it is obeyed and a promise is satisfied when it is kept” (1993: 49). The mental model associated with an utterance of the command “bring me of glass of water” will, then, include the mental representation of the desired act of the receiver bringing a glass of water to the speaker.

If instances of comic discourse can be represented as speech acts, then, it will be possible to suggest a theory of comic discourse in terms of the mental models that represent the conditions of their satisfaction. Raskin (1985), for one, briefly discusses considering jokes as speech acts, conceding that “[i]t is not too difficult to propose a definition of the speech act of making a joke” (1985: 328) but rejecting the approach as “reductionist” (Ibid.). Raskin’s concern seems to rely on the risk of giving too much importance to the illocutionary aspect of jokes, and humorous texts in general, at the expense of an understanding of their content. This legitimate concern, on the other hand, should not result into the opposite error of forgetting the importance, for the production of comic discourse, of the intention to obtain a specific reaction in the audience.

In the light of these requirements and concerns, the theoretical idiom of propositional attitudes and/or sentential attitudes, in combination with the notion of mental models, can be seen as even more promising for a theory of comic discourse than the idiom of speech acts itself. While understanding the satisfaction conditions of “bring me a glass of water” can be seen as relatively easy, defining the satisfaction conditions of an act of comic discourse is more problematic. The most natural suggestion would be to identify these conditions with the production of laughter in the audience; nevertheless, not only is laughter a very heterogeneous phenomenon (see Chafe 2007), but satisfaction conditions might include also smiling, frowning or, according to my experience, even recoiling in (a paradoxically amusing type of) horror. In other words, the complexity of the desired response associated with comic discourse would make every attempt to define it in purely behavioral terms as reductionist as Raskin feared.

The same, however, cannot be said with regard to a description in terms of propositional or, better (taking into consideration the importance, for some jokes, for instance puns, of their linguistic surface structure, not only of their meaning) sentential attitudes: comedy discourse, in these terms, can be seen as constituted by (among other things, but with specific importance) a set of sentences “S” (the punch lines) delivered with the intentions of producing in the recipient “X” a specific attitude towards them: the attitude “X finds S funny”. The task of understanding what this finding S funny sentential attitude actually means and entails, i. e. what people find funny and why, is probably the ultimate goal of the entire field of humor research. As such, in the present context this sentential attitude will need to be proposed as an undefined element of the theory here outlined.

Applying to this specific discourse situation the apparatus of mental model, the following dynamic model of what happens in oral self-translation of stand-up comedy can be proposed. First, the comedian/translator produces a mental model representing the sentences in the source language, their intended (and/or, as will soon be clarified, actual) audience and the finding S funny attitude of these audiences towards them. This is usually based on actual performances, so, as well as the semantic memory required to store the sentences themselves, this mental representation will also involve the episodic memory of reactions from specific audiences in specific performance situations. It is, however, also usually the case that the same sentences were delivered (although still in the same language) in front of different audiences and in different situations, typically with different results. While based on episodic memory of actual reactions, then, this mental model will arguably also require a certain level of generalization.

Secondly, the comedian/translator produces a mental model of a discourse situation in which the source language is replaced with the target language and the (remembered and/or generalized) source audience is replaced with the anticipated target audience. At this stage, for each relevant sentence, i. e. (at least) for each punch line, the comedian/translator produces a mental translation of the sentence into the target language and runs the model in order to predict whether each candidate translation is likely to produce the desired finding S funny sentential attitude towards it for the anticipated target audience. As Johnson-Laird (1983) suggested, the possibility of prediction is the greatest advantage offered by mental models. If the prediction does not produce the satisfied result in this mental simulation, for example because the cultural reference used in the punch line is likely to be unknown to the target audience, a better candidate translation is produced and mentally tested.

In the next phase, the translated text is tried out in front of an actual audience and the response is fed back into the model used for future translations, which at this point (just like the model for the source text) will also partially rely on memory and experience, not only on prediction. The translations that did not work will be revised and these modified versions will be played “in the theatre of the mind” in search for a better translation, until they are also tested in front of a real audience. If the new translations succeed in producing the desired response, they are kept, otherwise the cycle is repeated (the literature on plans calls this a Test-Operate-Test-Exit, or TOTE, plan. See Miller et al. 1960: 27), until either a successful translation is found, or the sentence is declared “untranslatable” and dropped from the target text.

Interestingly, the evidence collected from the interviews suggests that this mark of untranslatability was only applied to wordplay and to those punch lines based on culture-specific elements for which no functional equivalent was found, which were considered the exception. An example can be seen in the following excerpt from my interview with Romina Puma:

GP – Can you remember cases when the translation of material resulted particularly problematic?

RP – Not that many, most of my material works equally well in a language as in the other. Obviously there are some references that needed to be adapted.

All the participants interviewed seemed to agree on this relatively non-problematic nature of their self-translation. Moreover, some of the adaptations were perceived as more effective than the original ones, even leading to the retrofitting of the new reference to the original language when possible.

6 Conclusions

The notion that stand-up comedy performances are enabled by a mental text and the related notion of oral self-translation from a mental text bring with it a sense of opacity to direct observation that calls for a modelling effort in theoretical and sometimes abstract terms. In the previous sections, this effort resulted in the suggestion of a theoretical apparatus that comprises elements borrowed from the fields of cognitive science, psycholinguistics and philosophy of language. By using a phenomenological approach based on interviews, this theory was, where possible, corroborated against the experience of the translation activity collected from a sample of bilingual comedians. The interviews, indeed, both corroborate the idea that conscious translation does take place and that this process does not require the support of a written text. Moreover, they corroborate a conception of stand-up comedy in which some elements are constant and repeatable, while others are left to improvisation and variation. The result is an idea of the (mental) text of stand-up comedy based on a combination of declarative and procedural memory, with the declarative part considered essential to translation and constituted, most importantly, by the punch lines. Secondly, a dynamic model of its translation was suggested, based on the notion of a mental model representing the (predicted first, actual later) comic efficacy of these sentences. The hope, then, is that this double theoretical model might be useful in capturing, on one side, the complex nature of stand-up comedy and, on the other, the iterative and open-ended nature of its oral self-translation. More generally speaking, this effort can offer a contribution to filling a gap in our understanding of those forms of oral communication that, on one hand, do not represent instances of mere conversational spontaneity and, on the other, do not seem to be completely “scripted” either. Stand-up comedy, then, can still be investigated as a form of conversation, but this time as a conversation with a content, which can, inter alia, represents the object of translation and be investigated as such.