Blurring the Boundaries between Real Worlds, Discourse Worlds and Text Worlds Cases in Point: Kylo Ren and Black Widow’s ‘Marriage Story’, and Mark Wahlberg’s ‘Shooting’ of Matt Damon

,


environment. 'Black Widow Breaks Up With Kylo Ren' (internet source 1) and 'Black Widow and Kylo Ren Struggle With Their Relationship in
(internet source 2, capitalization as in original) are just two headlines that exemplify how the 'Marriage Story' movie tended to be pitched to audiences.
Simultaneously with the spring 2020 pandemic social distance restrictions, in the first half of June 2020 another socio-political event of international significance occurred. The 'Black Lives Matter' movement swept across the United States and, to varying degrees, across many European countries. At the time, I happened upon a Reddit post reading, 'The scene in The Departed where Mark Wahlberg shoots Matt Damon is the blueprint for how to handle corrupt cops' (Internet source 3).
Both the 'Marriage Story' and 'The Departed' references and interpretations cited above may be considered by some to be confusing or incoherent. To me, these references and interpretations are both interesting and worthy of analysis because they seem representative of specific present-day text-use and discourse-use tendencies. To support my point, let me clarify that 'Kylo Ren' is a character in the Star Wars franchise, and 'Black Widow' is a character in The Avengers franchise. 'Kylo Ren' is played by the actor Adam Driver, while 'Black Widow' is played by the actress Scarlett Johansson. In the Marriage Story movie, the two actors play 'Charlie Barber' and 'Nicole Barber' respectively. Neither character has anything to do with Star Wars and The Avengers.
In contrast, The Departed reference mentions Mark Wahlberg and Matt Damon, who are actors in this movie; however, it is, first, the same movie they participate in and not in different franchises as is the case with Adam Driver in Star Wars and Scarlet Johansson in The Avengers. Second, there is no reference at all to their characters. In other words, what interested me and made me combine the two movie references into a single paper is the fact that they both combine mentions of actors and movie characters but they seem to do so through different cognitive techniques.
Thus, the present paper addresses the question of how those two instances of movie-world references differ. The line of argumentation supported here suggests that to understand the nature of either of the examples and the difference between them, a World-building theory should be employed. Clearly, adopting a World-building theoretical approach would also require that the general theoretical background of our investigation is cognitive.
The data set used for the present purposes includes 50 movie screenplays, 50 political speeches, and 50 excerpts of virtual interaction (such as Twitter messages/Facebook comments/internet forums/social media comment sections, etc.). All the collected samples converge around two major features: (a) they exemplify public communication and public discourses; and (b) they exemplify the socially significant point of overlap -and blurring -where perceptions of public communication and personal communication meet. The method adopted -and, consequently, the analysis conducted here -is qualitative.
The paper is organized as follows: After this introductory section, the second section will present and discuss the general theoretical background against which the investigation is conducted. The third section will provide a discussion of the specific way in which the premises of the theoretical framework could be advanced in the present investigation. The fourth section will present analyses of data set samples with a view to demonstrating how the adopted theoretical approach actually applies in hands-on analysis. Then, the fourth section will return to the cases in point from the introductory section and will analyze them. The conclusion section will formulate generalizations and provide suggestions for the possible future applicability of the results of the present investigation.

World-building Theories
In understanding why World-building theories can be relevant to the present investigation, one first needs to take into account the fact that World-building theories stem from two different but not mutually exclusive sources.
Chronologically speaking, de Beaugrande and Dressler's ground-breaking work in the domain of Text Linguistics (de Beaugrande & Dressler, 1981) popularized the notion of the 'textual world' into scholarly language-and textoriented research. To de Beaugrande and Dressler, a Text World (henceforth TW) is associated with the notion of coherence; it is a well-organized configuration of cognitive structures which function together as a network and 'hold together' a whole text. Importantly, the emphasis in this formulation lies on the cognitive and non-language-bound nature of a TW. In de Beaugrande and Dressler's framework, language serves just to signal the procedural building of non-linguistic cognitive structures and networks such as a TW.
Furthermore, de Beaugrande and Dressler's TW 'contains' more than the cognitive structures signaled by the expressions in the surface text; it also 'contains' what de Beaugrande and Dressler view as associatively evoked commonsense knowledge structures, associatively evoked cognitive expectations, and the associatively evoked previous experiences of the participants in the communicative situation. In other words, de Beaugrande and Dressler's TW is an unambiguously mental construct which interprets reality. Crucially, de Beaugrande and Dressler's TW is not about 'true' or 'false' renditions of the real world (henceforth RW). Reality, in their viewpoint, is the conventionalized situation that is generally considered valid by a society or nation, and a TW may or may not cohere with this socially and culturally accepted version of what reality is.
Again, chronologically speaking, de Beaugrande and Dressler's understanding of TWs may seemingly have found its evolution in Werth's works (1999) andthrough Werth's works -into research in Text World Theory (e.g. Gavins, 2007;Whiteley, 2011) and Cognitive poetics (e.g. Hidalgo-Downing, 2004;Semino & Culpeper, 2002;Semino, 2010;Stockwell, 2002Stockwell, , 2009. Although published much later, most of Werth's works were actually written against the background of relatively the same scholarly environment as was de Beaugrande and Dressler's book. For instance, as Werth (1999, p. 20) acknowledges, he borrows the very term of TW from van Dijk (1977) -an influence also acknowledged by de Beaugrande and Dressler.
Generally, Werth's investigations focus predominantly on how different possible realities emerge in the process of text construction (these realities are believed to be associated with the use of negation, modality, etc.). Thus, Werth's proposal focuses on worlds and their subordinate sub-worlds, which he classifies into 'want worlds', 'belief worlds', 'intent worlds', 'desire sub-worlds', 'dream sub-worlds', etc. Moreover, Werth proposes that a TW should unavoidably "fall within the definition of 'mental space' of Fauconnier (1985)" (Werth, 1999, p. 20). However, even in the earliest inquiries into mental spaces, Fauconnier and his collaborators express their unwavering position that "[b]ecause the explicit structure set up in spaces is minimal, and because the default structure is always revisable as discourse unfolds, spaces are very different sorts of things than worlds (such as logicians' possible worlds, or the fictional worlds of narratives)" (Fauconnier & Sweetser, 1996, p. 12). In a similarly adamant way, Coulson also states that "mental spaces are local rather than global and are thus not describable in truth-conditional terms. This entails that the elements in a given mental space do not refer directly to entities in the outside world" (Coulson, 2001, p. 21). The present investigation subscribes to this latter treatment of Worlds as not related to classical logicians' views and interpretations of truth-objectivism. It is de Beaugrande and Dressler's interpretation that is preferred here due to its explicitly non-truth-conditional approach to the relation between TWs and RWs.
Nevertheless, there is a strong point in Werth's theory which is fundamental to the present investigation and which can still be harmonized with de Beaugrande and Dressler's interpretation of TWs: this is Werth's premise of the possibility that a TW can coincide with its corresponding discourse world (henceforth DW). A TW, from Werth's point of view, is a situation "distinct from the immediate one of the language event"; more precisely, it is "the 'story' which is the subject of the discourse, together with all the structure necessary to understand it" (Werth, 1999, p. 87). A TW, in his view, 'contains' people and objects as characters (Werth, 1999, p. 82), while a DW 'contains' participant roles (i.e. speaker and hearer). Crucially, it is quite possible for a DW and its corresponding TW to coincide, and such instances could freely be interpreted as "discourses about the discourse world itself" (Werth, 1999, p. 86). In cases like this, Werth maintains, a TW is "the conceptualization of that part of the discourse world which is 'in focus' for the purposes of discourse" (Werth, 1999, p. 86). This can be seen, for instance, in everyday conversations at home, in the street, at the office, etc. Such conversations concern interpersonal relations within a family, company, etc. Consequently, it is safe to argue, as Werth does, that participants in a DW can be constructed as characters in its corresponding TW. In such cases, the two types of Worlds overlap to a considerable degree, although a complete overlap or coincidence between them is not obligatory. This breakthrough notion of Werth's, coupled with de Beaugrande and Dressler's principles of World building, is the basis on which the present theoretical approach rests.

Embodied Realism
The fact that the present investigation specifically subscribes to de Beaugrande and Dressler's TW interpretation is conditioned by the general theoretical belief in Johnson's crucial realization (as acknowledged in Lakoff, 1987) that all human understanding depends on the nature of the human body, on the firing of neurons in our brains, and on our innate perceptual capacities and motor skills (for a discussion of the Neural Theory of Metaphor, see Lakoff, 2009). A claim such as this derives from the theory of embodied realism (Lakoff & Johnson, 1980, 1999, according to which our bodily experience (and not simply our social experience) is what gives rise to and then controls our linguistic and social functioning. In this theory, the body, through its inborn capacities and bio-physical characteristics, is what allows for the existence of specifically human cognitive abilities and, thus, it is also what enables human cognition: "every action our body performs is controlled by our brains, and every input from the external world is made sense of by our brains" (Lakoff, 2009, p. 1). Moreover, "[t]hought is physical" in that all mental structures are physically processed by brain structures, and the human body is so much 'within' the mind that "there is no aspect of our understanding that is independent of the nature of the human organism" (Lakoff & Johnson, 1999, p. 209).
Pivotal for such a standpoint is the human capacity for imagery (Johnson, 1987), i.e. for constructing alternative images of the same real-world phenomenon. 'Alternative' here needs to be seen, first, as crucial; second, it needs to be seen as ensuing from the premise that there is no one-to-one correct, objective, true, factual, etc. correspondence between anything 'external' to human thought and anything 'internal' or residing 'in' it. As a consequence, any human image, model, or (in our case) World is -by definition and by electro-neurophysiological necessity -a simulation of reality (or some part of it). In other words, human data processing is bound to be only interpretative -a phenomenon which associates with both the pervasiveness of the 'embodied mind' principle (Lakoff & Johnson, 1980, 1999Turner, 2011Turner, , 2014 and the 'embodied realism' doctrine (Lakoff & Johnson, 1999). This particular brand of 'embodied realism' is the kind of approach endorsed in the present investigation; it is the basis for our understanding of TWs, DWs and RWs.

Real Worlds and World Overlaps
Possible questions that could ensue from the above endorsement of Embodied Realism as a theoretical background to the present discussion may concern the presence and nature of RWs. As far as de Beaugrande and Dressler's and Werth's research on the issue is concerned, it does focus on TWs and on the coincidence of TWs with DWs, but it does not employ the notion of RWs. So, why do we need the notion of RW, and why are the notions of TW and DW insufficient?
The first possible answer is the existence of variations in the scope of a DW. To illustrate what is meant here by 'variations in scope', let us take the DW of a Parliamentary political speech. Such a speech may be delivered within the physical limits of a Parliament Chamber, but it may also be simultaneously broadcast to audiences far away from the actual physical environment of the person delivering it. Thus, to the person delivering it, the political speech's DW may only employ cognitive constructs of MPs as LISTENERS; alternatively, it may include cognitive constructs of MPs, SENIOR CITIZENS FROM LARGE CITIES, COUNTRYSIDE TV VIEWERS, YOUNG MOTHERS, etc. Furthermore, such a speech may be recorded and aired later, which would also introduce changes to its DW; in this case, the changes would not be in terms of location but of time. In brief, in such cases, the same text could evoke either a DW seen as encompassing cognitive constructs related exclusively to the communicative 'here and now', or a DW seen as encompassing a greater number of cognitive constructs that are not related exclusively to the communicative 'here and now'. The existence of these alternatives could be interpreted as the existence of a narrower or broader DW scope for the same text.
This principle is also revealed in variations in individual perceptions of a DW. As argued in the previous section, there cannot be expected to be a single, correct interpretation of (the scope of) a DW that is valid for all members of an audience. Moreover, it is highly questionable if a single, correct DW interpretation could exist even for one and the same text receiver: today, that text receiver may interpret a DW rather differently from the way they interpreted it, for example, two weeks ago. Their interpretation could be expected to vary, if in the meantime, for instance, the political speech deliverer realized their speech was recorded and aired after its original delivery.
This possibility for information about the actual, i.e. real, communicative environment to change one's perception of a DW is another issue that needs to be addressed. In essence, the issue concerns the interplay between a DW and reality. The first question here is how far a DW can extend. The second and (in my opinion) more important question is where does a DW 'extend'? In other words, what exactly is the space that a DW can draw on, extend into and 'engulf' chunks from? The answer necessarily has to underline the fact that that space can only be a conceptual space due to the fact that a DW itself is one. As our example with the political speech demonstrates, a DW can incorporate concepts related to a broader and physically distant reality, but it cannot directly incorporate people from that reality, hence the present choice to postulate an RW as a conceptual space which interprets reality but is not equal to that reality.
Three main points seem to need clarification regarding my interpretation of the difference between an RW and 'reality'. The first point is that my interpretation relies on the principles of cognition, therefore it does not approach 'reality' as referring to a fully describable and fully analyzable phenomenon which stands independent of human understanding and needs to be fully matched or perfectly represented by human cognitive structures and human academic theories. My interpretation approaches 'reality' simply as a describable and analyzable phenomenon (though not fully so) which does stand outside human minds and of which human cognition can (and needs to) create simulations. In turn, an RW is the human cognitive interpretation/simulation of what 'reality' may be. A perfect fit between the two, I believe, is not needed as humans can and do manage to function on no more than 'sufficient' fits between reality and their RWs.
This discrepancy between a viewpoint on 'reality' as fully amenable to human representation and 'reality' as no more than something sufficiently well simulated by human cognition can trigger a significant blur of the theoretical and analytical focus. What can remain outside our lens is the important assumption that human information processing and exchange, of which textual communication is part, might not be dealing directly with an objective reality. The critical shift of angle is that textual communication -like any other human activity -has to do directly only with mental simulations of an objective reality and not directly with reality itself. Human-specific information processing mechanisms invariably stand as the mediator between the two; it is my conviction that a theoretically sound approach could only benefit from taking this into account.
Such a claim, it should be highlighted, does not represent a denial of the existence of an objective reality. The argument here holds that it is the specificities of the human brain and our sensorimotor pre-conditioning (Lakoff & Johnson, 1999) that make it impossible for humans to understand and interact with a (possibly) objective reality. As already argued in the Section above, without the specificities of the human brain and our sensorimotor pre-conditioning, we, as humans, would not be able to experience reality in the particular ways we do. As a result, our cognitive abilities, which arise from the specificities of the human brain and our sensorimotor systems, unavoidably mediate between the (possibly) objective reality and our human perceptions of it. In a quite significant way, reaching beyond our cognitive abilities and directly into reality seems simply not possible for humans (Lakoff & Johnson, 1999). What seems much more attainable for linguists, as humans, is the study of the human version, i.e. the human interpretation, of the (possibly) objective reality. No matter how opaque, principles of human mental modelling and simulation still appear to be more accessible for us than the reality they serve to interpret. Consequently, analyses of reality are also more likely to be attainable if they are approached as studies of human cognitive structures and not as studies of direct interaction with an objective, human-independent reality.
Perceptions of objectivity, on their part, are most likely brought into effect through high frequencies of occurrence of similar human-nature-conditioned mental interpretations. Due to the common build of human neurological and sensi-motor systems, human-produced mental structures tend to seem significantly similar to each other, and this similarity can be taken to bring about the impression of objectivity, or what Johnson (1987, p. 18) calls 'one correct God's-eye-view'.
In sum, the present paper upholds the possibility -or, indeed, the necessity -to approach reality from the perspective of the cognitive mechanisms employed by the participants in a communicative situation and not vice versa. In other words, the perspective employed starts from the human mind; then, it moves towards the possible reality that the human mind needs to interpret/ simulate. Even more simply put, what I uphold here is that the generally accepted viewpoint on textual analysis should be reversed and the cognitive structures and mechanisms used by communication participants should not be simply added to the object of analysis: it is the context which should be seen as part of the cognitive structures and procedures processed by the participants. After all, reality exists for humans only because they see it as such. An objective reality may or may not exist outside human perception, and no 'World' needs to be seen as a special case of 'real-ness'. In a way -and this can be argued to be quite a crucial way -we can study only our human-specific perceptions of reality and not directly that reality itself.
As a consequence -and this is the second point that seems to need clarification -what is perceived here to be an RW is just one of the numerous possible interpretations of reality. An RW is seen here as an interpretation that need not be an accurate one but needs to be a sufficiently adequate simulation of reality which allows a person to function sufficiently adequately in said reality. Furthermore, that person's current RW need not be a perfect fit to reality, not even to the same person's reality of, let's say, today or two years before today. RWs can vary and this, I believe, is confirmed by everyone's experience of 'learning'.
The third point that seems to need at least a brief clarification is the one concerning the conceptual 'relationship' between an RW and a DW. As I have argued elsewhere (Tincheva, 2018), from a cognitive perspective a DW can be seen as standing metonymically (on conceptual metonymy see, e.g., Barcelona, 2000Barcelona, , 2011Dirven & Pörings, 2002; Ruiz de Mendoza, 2020) against the background of an RW. A DW can draw on an RW and it can do so dynamically (Tincheva, 2018), by either narrowing or expanding its scope.
As our example of a political speech shows, a DW may 'overlap' with and 'extend' into an RW to various degrees. This, however, does not cancel the cognitive necessity for a DW to cohere sufficiently with an RW (for a cognitive interpretation of 'sufficiently', see de Beaugrande and Dressler's discussion (1981) of cognitive THRESHOLDS). In other words, even if a text producer is lying, they will need to present a DW as overlapping with its RW, no matter if in their mind the two actually diverge. Importantly, the same principle applies not only to DWs. TWs also need to cohere with an RW, even when they are far removed from it. A novel, for instance, can select a portion of life and so construct it as to purport realism. Conversely, a sci-fi book could configure a representation of what life is expected to be or what it should be. The resulting TW would be just a simulation of a possible RW. In other words, adequacy of representation, or a fine fit between a TW and an RW, is not an obligatory requirement; this is also evident, for instance, in fairy tales and Hollywood movie scripts. The purpose of creating a TW and a DW, therefore, can be defined as constructing a coherent and convincing albeit not necessarily realistic interpretation of an RW. This RW, however, would unavoidably function as the default background against which any TW or DW would be interpreted.
Understanding the relationship between a figure and its background entails understanding the cognitive phenomena of profiling and profile shifts. Generally, these notions derive from psychological studies of visual perception (especially Rubin, 1915Rubin, , 1958 and of gestalt principles (Koffka, 1935;Wertheimer, 1923). The basic premise here starts from the innate inability of the human mind to process information flow in terms of infinite homogeneity. Imposing hierarchical orders, to put it more aptly, is practically indispensable in any human thought process as some information elements need to be selected and brought into focus, while others need to be 'pushed' to the background. As is evident from the so-called Rubin's vase (Figure 1 below), one needs to choose whether to look at the white portion of the picture or the black one. Seeing them both simultaneously is not a possibility. Moreover, profiling is not a static cognitive phenomenon. It is a dynamic cognitive technique in that what serves as the background at one moment could become profiled at the very next one, and the prominent figure of the previous moment would shift into the background position. The analysis below tries to extend this dynamic view of profiling to the domain of operation of RW, TW and DW.

Sample Analyses
As stated in the introductory section above, the data set used for present purposes includes 50 movie screenplays, 50 political speeches, and 50 excerpts of virtual interaction (i.e. Twitter messages, Facebook comments, internet forums, social media comment sections, etc.). First, this section presents analyses of data set samples as a demonstration of how the theoretical approach adopted here applies to hands-on analysis. Second, the section returns to the two cases in point from the introduction and offers their analysis, too.
In exemplifying the construction of a TW, a representative sample from our data set would read as follows:

(Sample 1)
From the moment that the French defenses at Sedan and on the Meuse were broken at the end of the second week of May, only a rapid retreat to Amiens and the south could have saved the British and French Armies who had entered Belgium at the appeal of the Belgian King; but this strategic fact was not immediately realized. The French High Command hoped they would be able to close the gap, and the Armies of the north were under their orders. Moreover, a retirement of this kind would have involved almost certainly the destruction of the fine Belgian Army of over 20 divisions and the abandonment of the whole of Belgium. Therefore, when the force and scope of the German penetration were realized and when a new French Generalissimo, General Weygand, assumed command in place of General Gamelin, an effort was made by the French and British Armies in Belgium to keep on holding the right hand of the Belgians and to give their own right hand to a newly created French Army which was to have advanced across the Somme in great strength to grasp it. (Winston Churchill, PM;March 5, 1946) No linguistic structure in this longish passage serves to evoke a DW. The excerpt focuses exclusively on creating a TW. From the moment and at the end of the second week of May evoke the mental construction of past moments and periods and not of the time of the current communicative process, thus providing temporal specifics for the TW only. The spatial delimitations of the TW are specified through the use of, e.g., Sedan, the Meuse, Amiens and the south, Belgium, the Somme. Moreover, all cognitive structures connected with PERSONS and evoked by the text are of people not present at the time and place of delivery of the speech: e.g. THE BRITISH AND FRENCH ARMIES, THE BELGIAN KING, THE FRENCH HIGH COMMAND, FRENCH GEN-ERALISSIMO, GENERAL WEYGAND, GENERAL GAMELIN, THE BEL-GIANS. Consequently, no element in this passage can be considered as bringing about the construction of a DW.
Similarly, Sample 2 below also evokes no concepts associated with what could be interpreted as a DW. The whole passage focuses on the creation of a TW through the coordination of concepts such as THE CANARY WHARF BOMB, THE REPUBLICAN MOVEMENT, THE BRITISH AND IRISH GOVERNMENTS, TERRORISTS, THE LAW, CEASEFIRE, etc. All these concepts represent entities absent from the immediate communicative environment:

(Sample 2)
The Canary Wharf bomb clearly and conclusively proves that there is no prospect of the republican movement becoming committed to exclusively peaceful means.
The duty of government is clear. But the British and Irish Governments have shirked that duty. Instead of making terrorists amenable to the law, they have responded to their agenda asking in return only for a "credible ceasefire". (Trimble, 23 March 1996, UUAP Annual General Meeting) An illustration of the existence of overlap between a TW and a DW can be found in Sample 3. This dialogue exemplifies how conceptual structures referring to PARTICIPANTS in a DW can simultaneously function as CHAR-ACTERS in the TW:

(Sample 3)
Wife:…and the kid is running a temperature. Husband: He is? Wife: Probably a tooth or something. Husband: Let's hope it's just a tooth. We'll have to watch closely… Wife: You mean, I will. Husband: Yeah, you will. If you want, we can switch -you go to work, I'll take care of him… In this Sample, us, we, you, and I refer both to the participants in the communicative exchange (i.e. DW participants) and to the characters in the TW, i.e. to the HUSBAND and WIFE. Therefore, us, we, you and I trigger the TW and the DW simultaneously. The he reference, in contrast, is dubious. If the child were not present in the immediate communicative exchange, the reference of he would point to the TW only; if the child were present, he would trigger both the TW and the DW. This ambiguity is also one of the confirmations of the existence of different degrees of overlap between TWs and DWs. In other words, the use of he, if triggering both the TW and the DW, will add to the conflating force of us, we, you and I, thus increasing the extent of overlap. Consequently, the TW-DW overlap, or coincidence, is gradable in addition to not being obligatory.
The presence of the TW-DW overlap is not characteristic only of nonpublic types of discourse (as in Sample 3). Despite the fact that both Sample 1 and Sample 2 above do not reveal any degree of World coincidence, public political speeches can also display TW and DW overlapping. This possibility is employed, for example, in President Bill Clinton's televised White House Speech of December 11, 1998, in which the cognitive construct of the leader in the TW and the cognitive construct of the speaker in the DW coincide, while the citizens in the TW coincide with the listeners in the DW:

(Sample 4)
As anyone close to me knows, for months I have been grappling with how best to reconcile myself to the American people, to acknowledge my own wrongdoing and still to maintain my focus on the work of the presidency.
Sample 5 below employs the same cognitive techniques but, in contrast to Sample 4, it is also representative of how profile shifts can function in parallel with World overlaps.

(Sample 5)
MY LORDS AND MEMBERS OF THE HOUSE OF COMMONS. My Government's central objectives are economic stability, and investment and reform in public services, leading to a more prosperous and inclusive society. To this end, my Government will continue to secure low inflation and sound public finances. My Government's main priorities for the forthcoming Session will be reform in education, health, crime and welfare. (Queen Elizabeth II;20 June 2001) In Sample 5, My Lords and Members of the House of Commons directly triggers the activation of PARTICIPANTS in the DW, thus profiling the DW and suggesting that the TW should be pushed to the background (at least those structures of it which do not overlap with DW structures). Then, the remaining part of the text focuses on the construction of a TW. However, every time my (in my government) is used, it first signals that both the DW and the TW are activated. Second, it indicates which of the overlapping World networks should be brought to the fore. If all the government are present at the time of speech delivery, then what signals the need to switch from profiling the DW to foregrounding the TW is the end of the whole noun phrase my Government.
If the government are not present at the time of speech delivery, then only the possessive pronoun would be used to signal the DW. In such a case, what signals the need to switch from profiling the DW to foregrounding the TW is the demarcation line between the pre-modifier and the head noun within the noun phrase my Government. Similarly, central objectives are economic stability, and investment and reform in public services, leading to a more prosperous and inclusive society suggests the TW be pulled to the fore and the DW be suppressed as the background.
To support the claim that the principle of profiling holds for all three types of Worlds, let us consider the following sample from the How to Train Your Dragon movie screenplay. The text opens with a voice-over (i.e. an off-camera commentary) directly addressed at the text receiver: (Sample 6) HICCUP (V.O.) This is Berk. It's twelve days North of hopeless, and a few degrees South of freezing to death. It's located solidly on the meridian of misery. My village. In a word, sturdy. And it's been here for seven generations, but every single building is new. We have fishing, hunting, and a charming view of the sunsets. The only problems are the pests. You see, most places have mice or mosquitos. We have … dragons.
In Sample 6, the village of Berk is constructed as the place of 'here and now', as required by the deictic demonstrative this. But does this 'here and now' belong within the DW, as the voice-over technique suggests? Does it belong to the RW as a four-year-old would believe? Or does it belong to the TW? Furthermore, if BERK is interpreted as part of the DW, then whose 'here and now' does that DW represent? Is it the speaker's, and who, indeed, is the real speaker in that text? Is it the script writer or is it the main character (i.e. Hiccup)? Furthermore, whose opinion should be valid on these questions: is it the film makers', the actors', the film critics', or the opinion of the children who watch the movie and who are, presumably, the intended audience of the text-presented story?
The answers to these questions would probably depend on the answer to another one: the question of who the intended text receiver is. Is it the film producer, who first reads the script and on whose decision the potential making of the script into a film depends, or is the intended text receiver, namely the future film audience? Furthermore, do the speaker and the text receiver share the same 'here and now', when, by default, the text producer (i.e. script writer) and text receiver (i.e. movie producer or his/ her assistant) rarely share the same space (i.e. they are rarely present together at a film producer's office or home, where the movie script is first to be read and, then, turned into a screenplay)?
Clearly, the background knowledge of a normally socialized adult would suggest that, in the movie genre, the text producer is the script writer 'speaking' 'through the voice' of the main character. However, for this to happen, a person needs to have accumulated enough conceptual experience to also understand that the TW (the world of HICCUP) is not really a part of the RW, although it is presented as such (i.e. the two Worlds do not really overlap). An adult would know that the text receiver's DW will necessarily be different from the main character's DW, which, although constructed in the form of a DW, is still nothing but a part of the fictive TW. An adult would also have previous knowledge of the fact that, in this way, the TW of the main character is manipulatively constructed as (if) a part of the RW and also as a DW that is meant to be seen as 'shared' with any future text receiver. And last but not least, Sample 6 also illustrates that it is not only the TW and the DW which are profiled against the RW. As the use of You see, most places have mice or mosquitos reveals, sometimes even the RW can be pulled to the fore, and the TW and the DW can be suppressed.
What is of greatest significance here, however, is the mere fact that the above questions and ambiguities exist. What prompts these questions and gives rise to the ambiguities is the existence of a blur between the DW, the TW and the RW of a text. If the three types of Worlds were not all co-activated, their 'overlap' would not be possible, and neither would be the above questions and ambiguities. Once co-activated, the three Worlds function through a number of cognitive mechanisms, a major part of which are profiling and shifts in profiling.
Applying all these precepts and thus analyzing the two cases in point from the introductory section, one finds that in the 'Black Widow Breaks Up With Kylo Ren' article title, Black Widow simultaneously signals a TW (through the person the article's 'story' is about) and an RW (through the cultural entity of a Marvel comics character, or as many would have it, a Marvel 'Universe' hero). In other words, Black Widow signals a blending or blur between a TW and an RW. Kylo Ren also signals such a blending or blur between a TW and an RW; the only difference is that this particular character associates with the Star Wars RW cultural domain and not with the Marvel one. Importantly, the Marvel-based TW and the Star Wars-based TW have nothing in common, i.e. they do not overlap and thus represent two separate TWs evoked by the same article title.
Breaks Up in the article title, however, represents quite a different mechanism. It evokes the TW of the movie that is the main topic of the title, and that movie's TW has nothing in common with, i.e. it does not overlap with, the Marvel-based TW or the Star Wars-based TW. It is a completely new, third TW, and, as the title suggests, in that third TW, Black Widow and Kylo Ren were married and now are getting divorced. However, true to fact, this has nothing to do with what the third TW is about. The third TW is about the Charlie Barber and Nicole Barber characters, who are the ones actually divorcing. What makes this overlap of three TWs possible is the RW fact that the RW person (i.e. Scarlett Johansson) who plays the Black Widow character in the first TW is the same person who plays the Nicole Barber character in the third TW. Similarly, the RW person (i.e. Adam Driver) who plays the Kylo Ren character in the second TW is the same person who plays the Charlie Barber character in the third TW. In sum, the cognitive mechanism at work in the case of this article title allows for an RW to overlap with a total of three TWs, while there are also separate pairings and overlaps among these TWs.
In the case of the Twitter message 'The scene in The Departed where Mark Wahlberg shoots Matt Damon is the blueprint for how to handle corrupt cops', Mark Wahlberg and Matt Damon are both RW actors, and, as such, their mentioning evokes the construction of an RW. However, in the RW, Mark Wahlberg has not shot Matt Damon. Thus, shoots evokes a TW in which that happens. As the RW and this TW share structures (the concepts of MARK WAHLBERG and MATT DAMON), there is an overlap between the two. The Departed and The scene in The Departed also evoke an RW (one related to movies and movie structure), while handle corrupt cops evokes either a second RW or a different portion/domain of the first RW. In contrast to the Black Widow and Kylo Ren example, this case presents us with only one TW that coincides with one or possibly two RWs.

Conclusion
This investigation addresses the issue of how Discourse Worlds and Text Worlds can be created simultaneously through the same text so that the Worlds' boundaries become blurred and difficult to delineate.
The line of argumentation allows us, first, to contribute by providing a theoretical advancement to World-building approaches by supporting the necessity for the concept of Real Worlds to be introduced into related theories. Data set sample analyses employing the Real World concept are also included to verify the main theoretical premise. The texts analyzed associate with both traditional genres, such as political speeches, as well as with modern-day boundary-blurring genres, such as Twitter messages.
The analyses also confirm that the conflation of Worlds happens through the cognitive mechanisms of overlapping, profiling and figure-background profiling shifts. Thus, the analyses can be seen as successfully extending the dynamic view of these cognitive mechanisms to the domain of RW, TW and DW operation.

Blurring the Boundaries between Real Worlds, Discourse Worlds and Text Worlds
What do the article title 'Black Widow Breaks Up With Kylo Ren' and a Twitter message saying 'The scene in The Departed where Mark Wahlberg shoots Matt Damon is the blueprint for how to handle corrupt cops' have in common? Clearly, they both combine references to actors and movies, but do they combine them through the same cognitive technique(s)? This paper starts by addressing the question of how these two instances of reference differ. The line of argumentation that is supported suggests that a world-building theory needs to be employed in order to understand the difference. In doing this, the paper aims and contributes to the theoretical advancement of world-building approaches by arguing for the introduction of the concept of 'Real Worlds' in research on Text Worlds and Discourse Worlds. Data set sample analyses employing the Real World concept are included to verify the main theoretical premise. The analyzed texts cover traditional genres such as political speeches as well as modern-day, boundary-blurring genres such as Twitter messages.
Keywords: text world, discourse world, real world, cognitive linguistics, figure-ground profiling, Twitter messages