Narrative Ambiguity in Contemporary Cinema:

Narrative Ambiguity

in Contemporary Cinema:

A Cognitive Account of its Formal Strategies and

Relation to Unreliability

Name of the student: Nicolás Medina Marañón. Student number: S3796078.

Master’s degree programme – specialisation: Arts, Culture and Media – Film and Contemporary Audiovisual Media.

Title of final-year thesis: Narrative Ambiguity in Contemporary Cinema: A Cognitive Account of its Formal Strategies and Relation to Unreliability.

Name of thesis supervisor: Dr. Miklós Kiss (first supervisor) and Prof. Dr. Annie van den Oever (second supervisor).

I hereby declare unequivocally that the thesis submitted by me is based on my own work and is the product of independent academic research. I declare that I have not used the ideas and formulations of others without stating their sources, that I have not used translations or paraphrases of texts written by others as part of my own argumentation, and that I have not submitted the text of this thesis or a similar text for assignments in other course units.

Date: 5 May 2019. Place: Groningen. Signature of student:


Table of Contents

1. Introduction

1

2. Ambiguity and Unreliability

3

3. Individual Differences and the Arts

13

3.1 Personality Traits 13

3.2 Aesthetic Empirical Research 17

4. Examining Contemporary Ambiguous Films

23

4.1 Open Endings 23

4.2 Formal Features of Contemporary Ambiguous Films 28

4.2.1 Contradictory Hypotheses 28

4.2.2 Enigmatic Dialogues 43

4.2.3 Low Communicativeness of Main Characters 49

4.2.4 Scarce Use of (Non-diegetic) Film Music 59

4.2.5 Film Genre and Ambiguity 71

4.2.6 Meta-level Reflections and Unreliability 75

5. ‘Risks’ and ‘Rewards’ of Contemporary Ambiguous Films

82

6. References

89

7. Summary

100

1. Introduction

Ambiguity in fiction film, that is, the use of syuzhets that afford multiple, and in some cases mutually exclusive, readings of the fabula, has been present for many decades in the history of film, mostly permeating art-cinema narratives. However, in the last couple of decades, from the early 2000s onwards, the introduction of ambiguity in narrative films has found a place also in more (post)classical productions, leading to a progressive mainstreamification if not (sub)genrefication of ambiguous narrative cinema.

This current might be the result of a broader attempt to inject cinema with a breath of fresh air, similar to the varied narrative devices deployed by the trend of cinematic complexification, such as nonlinear narrations, multiplicity of ontological structures, doppelgängers, etc. However, while this general trend of contemporary cinematic complexity has received considerable theoretical attention, the more elusive (sub)trend of cinematic ambiguity and its recent proliferation seems to lack a similar amount of scholarly reflection. Therefore, my aim with the present text is to tap into this research gap with the intention to establish some preliminary examinations of the commonalities present in these contemporary ambiguous films — regarding their formal dimensions, general features, as well as their patterns of reception —, and by that attempt to substantiate my hunch regarding the existence of the ‘ambiguity-trend’ in contemporary cinema.

In the movies I will be examining, viewers are usually invited (at times forced) to refrain from obtaining a narrative closure on the state of affairs they are presented with. This is achieved via a series of strategies: the presentation of several (at times mutually exclusive) plausible interpretations of the fabula events; enigmatic dialogues thematising the notions of representation, illusion and reality; main characters presenting a low degree of communicativeness; sparse use of non-diegetic film music; a combination of met and unmet genre expectations; and lastly, indirectly prompting the viewers to engage on reflections on a meta-level about the representation act and the filmic artefact, from where the unreliability of the narration as a whole may emerge as well. With these strategies, these narratives tease the viewers by testing some of their personality traits such as their tolerance of ambiguity, tolerance of uncertainty, negative capability or need for cognitive closure. All of them determining factors in the way audiences will approach and deal with these films.

In a similar manner, and under the consideration which regards the arts as a representation and a reaction to the cultural context in which they are conceived, these films

are emerging in a Zeitgeist largely characterised by an increasing degree of unreliability in media representations. This unreliability is resulting from the inability people frequently face to obtain certainty in the assessment of varied information. While this uncertainty has always been a part of social life to some degree, new technological developments are heightening this concern, which seems to be pointing to an environment in which the individual’s distrust will be the default attitude. In this way, the unreliability resulting from the ambiguous narrations in the filmic context, seems to parallel the ever-growing unreliability experienced in our extra-diegetic climate.

Nonetheless, a very important distinction exists among these two cases: the ambiguity of the filmic narration can serve as attempts to dismantle or expand certain narrative paradigms and promote some sort of critical/optional thinking during these processes of aesthetic experience. Conversely, the unreliability populating the media atmosphere does not seem to work in the same way, rather, this unreliability is weaponized to enhance uncertainty about particular issues, and, instead of inducing any kind of reflection or reassessing of people’s worldviews, it is present here in the hope that it will reinforce already existing ideas, opinions and explanatory frameworks.

On the other hand, there are certain similarities between both experiences, particularly in relation to the notion of unreliable information. Both the experience of watching a film characterised by its ambiguous plot and the consumption of news stories (or similar media in which factuality is lessened) are, I hypothesise, giving prominence to the emotional and affective dimension of the communicative episode. Needless to say, of course, in spite of this resemblance, truth should not be weighing equally in both instances, as the effects for society in the case of news information are much more direct in comparison with the more aesthetically focused occurrence of the filmic experience. Emotions such as interest (or the lack thereof), curiosity, confusion, the degree to which individuals are able (prone) to use a simple/complex cognitive structure or the use of stereotypical classifications (master schemata) all carry an important role on both polarizing ideologies as well as in the process of coping with ambiguous plots.

Before going in depth into the different paths opened in this brief introduction, I will outline the structure of the present text. Initially, I will trace an overview of the characterizations and definitions of ambiguity in the context of the arts, and particularly within film narration, together with brief distinctions of closely related concepts such as vagueness, ambivalence or openness. Following this, I will take a deeper look at its relation to

unreliability, necessitated by the broader context of our Zeitgeist in which these concepts seem to function jointly. Next, I will proceed with a conceptual overview of the personality traits aforementioned and examine the ways they have been studied in relation to aesthetic experience. Subsequently, a description of Bordwell’s and Branigan’s models of narration will be presented in order to provide a richer ground for the subsequent close readings of the case studies. Then, the strategies employed by this trend of contemporary ambiguous films will be presented in combination with close-reading segments from each one of the case studies, namely Burning (Lee Chang-Dong, 2018), Certified Copy (Abbas Kiarostami, 2010) and Sunset (László Nemes, 2018), including references to films like You Were Never Really Here (Lynne Ramsay, 2017) or Synecdoche, New York (Charlie Kaufman, 2008), as well as others that will be mentioned in passing. Needless to say, these strategies by no means intend to constitute an exhausting categorical classification of the phenomenon, but a preliminary attempt at examining and by that verifying what I deem is a trend in contemporary cinema.

2. Ambiguity and Unreliability

As mentioned above, ambiguity as a general concept has been an object of study for a long time, presenting differences in its appreciation depending on the context. In this line, scholar Murray Smith points out that describing a proposition as ambiguous is an attribute that will easily evoke terror in the philosopher, while, conversely, it is aptly used as a term of praise in the case of art (Smith 2006: 40). Particularly praiseworthy in the artistic situation is the paradox this type of artworks entail, that is, the unification or holding-in-balance of contrasting attitudes or meanings (ibid.).

Taking the dictionary definition for ambiguity as a starting point — that is, “uncertain, open to more than one interpretation, of doubtful position” — neurobiologist Semir Zeki proposes a refinement from a neurological perspective (Zeki 2004: 173). He claims that true ambiguity appears when none of the proposed solutions is more likely than the others, which inevitable forces the brain to adopt the only option possible: considering all the solutions equally likely, and providing a place for each of them on the conscious stage, one at a time, which means being conscious (aware) of only one of these interpretations at any given moment (ibid.: 175). His descriptions of the meaning-instilling processes of the brain goes into depths beyond the scope of this text (contemplating notions of essential nodes,

micro-consciousness or conscious correlate), but, for our interest, the main idea to be extracted from this is the predisposition of the brain to imbue meaning into the world, however confusing the signals it receives are, with the final aim to finding a solution. Therefore, when the brain faces a situation where this obtaining of a solution is prevented, due to its confrontation with several meanings of equal validity, the only option is allowing the various interpretations (ibid.: 188). Hence, Zeki adjusts the aforementioned definition by suggesting that ambiguity is not so much characterised by uncertainty (like the dictionary proposes) but, “rather, the certainty of different scenarios each one of which has equal validity with the others” (ibid.: 189), that is, the certainty of the uncertainty. Moreover, and of particular interest for our focus on cinematic ambiguity, he distinguishes between graded steps in relation to the level of ambiguity and the number of areas affected by these stimuli. Places at the highest level are those ambiguous states involving distinct areas (cortical sites) that can exert their (top-down) influence: experience, memory and learning among other factors (ibid.: 192), which play a key role in the process of narrative comprehension.

As an example of an artistic stimulus of this nature, Zeki talks about the “genius of Vermeer,” in that by managing to convey a full array of excluding expressions (in The Pearl Earring), he does not provide a clear answer, but multiple ones. Each of which the viewer is only conscious of one at a time (ibid.: 189). Moreover, regarding the praised quality of ambiguity in art, Zeki relates this to the physiology of the experience and the capacity of the brain to provide several interpretations, a vital capacity for it to acquire knowledge (ibid.: 190). Accordingly, it is not the ambiguity itself that is pleasing but, rather, our brain’s capacity to experience this multiplicity (albeit we are conscious of only one at any given moment) in response to such stimuli (ibid.: 192).

From the different disciplinary position of philosophy, drawing from (and elaborating upon) the understanding of linguistic ambiguity, philosopher Isreal Scheffler studied the distinct phenomenon of pictorial ambiguity based on this same premise of multiple possible meanings. In doing this, he established a useful separation between vagueness and ambiguity: where the former would refer to the indecision of applicability of meaning, while the latter would present no doubts about the applicability of a meaning but, rather, a conflict among several satisfactory resolutions (Scheffler 1989: 110). Similarly, film scholar Lutz Koepnick draws a distinction between another pairing at times used interchangeably, that of ambiguity and ambivalence. Koepnick explains that while the former belongs to the aesthetic object, the latter is more strongly connected to the reception (evaluation, interpretation and validation) of

said object (Koepnick 2018: 107). That is, a text is to be considered ambiguous when its content points in different directions at once, ultimately refusing unequivocal resolution; conversely, Koepnick suggests, ambivalence is to be understood as a conscious/unconscious impulse that leads us to understand something in more than one way, a matter of “perception, psychology and phenomenology” (ibid.). However, as Koepnick himself recognises, this distinction — unlike that of vagueness and ambiguity — can be a very arduous one to establish, particularly if we consider the intrinsic relation between the object (film) ontology and its relation to the viewer. This entangled distinction is illustrative of the blurry territory upon which ambiguity is created, to which both the formal features of the text as well as the perception of such text (in the sense of the individual’s activity) contribute.

Returning to Scheffer’s conceptualization, ambiguity, for him, consists in the oscillation between rival interpretations, inviting to grasp and preserve this multiplicity rather than opting for an individual solution. Scheffler goes on then to speculate about the possible reason for the fascination ambiguity might evoke, positing that it is “the suspicion of inconsistency, the fear (or perhaps the hope) that logic has been breached” (ibid.: 114). That is, certain fascination would arise in perceiving ambiguous artworks, specifically due to the disruption of clear-cut interpretations these ambiguous stimuli and information signals could entail.

Berndt and Kopenick, in the introduction to their recent volume on ambiguity in contemporary art, while praising Scheffer’s contribution in regard to its addition to previous classifications of interpretive uncertainty (Berndt & Koepnick 2018: 8), they also identify what they consider to be the shortcomings of Scheffer’s account: its limitation to art forms based on language and its primarily classificatory nature; which they deem “offers no real lenses to inspire intricate readings of individual practices and materials” (ibid.). Conversely, they promote focussing their attention on ambiguity’s potential to unsettle stable expectations and to push against closure and certainty, while stressing the creativity of the act of reception, the prevailing of the inventiveness of questions over the authority of final answers (ibid.: 13). Viewed this way, then, when our meaning-instilling processes fail to orient us (in film narratives or any other aesthetic experiences) to arrive at certainty and clarity, this would not really be a failure, but part of the intended experience. Opening up, in this manner, new zones of aesthetic experience within the context of indetermination and multiplicity.

One of the most recent attempts at conceptualizing ambiguity in the arts has been carried out by art historian Verena Krieger. She first traces an overview of some key

contributions to this field by several theorists: William Empson and his Seven Types of Ambiguity (1949); Bakhtin’s concept of dialogism and his contrasting it with authoritarianism and monological qualities; or Kriss and Kaplan’s aesthetic ambiguity and their emphasising of its psychological function (Krieger 2018: 77-78). Then, she goes on to distinguish five levels of aesthetic ambiguity: the media level, concerned with the forms of ambiguity resulting from the specificities of the particular medium; the artistic level, ambiguity as a defining characteristic of art; the intentional level, interested in the instances in which the means of conveying meaning are suppressed deliberately, beyond the limitations of the medium and/or its artistic considerations; the historical level, which she considers the specificity of historic periods in relation to aesthetic ambiguity; and the receptive level, which points out the non-exclusivity of the structural dimensions of the artwork as the responsible for its ambiguity, stressing the process of reception as equally important, as “ambiguity always unfolds in the relation between an aesthetic object and a viewing subject” (ibid.: 85). These different levels will prove useful in the following pages when examining, for example, the formal dimensions of the case studies (intentional level), the meta-reflections about the filmic artefact and its parallels with media communication unreliability in general (media level).

Beyond this distinction in levels, and focusing on the structural aspects of the text, Krieger proposes a systematic perspective on aesthetic ambiguity, outlining four modes emerging from the combination of two parameters assessing the elements of information contained in the artwork, namely, the relationship of the information and the intensity of its representation. The former oscillates from working together to working against each other, and the latter from weakly manifested to strongly manifested (ibid.: 100). So, the first mode described by Krieger is the conjuctive mode, in which pieces of strong information work together, in this case the “chain of reactions created by the ambiguities on several levels constitutes a coherent whole” (ibid.: 98), presenting a tendency to superordinate explicitness. Next, when the information still works together but is weakly manifested Krieger talks of the associative mode, where it is possible to link associations and create meaning, but these connections are not conclusive enough, leading to vagueness as a result (ibid.: 99). Next, the indifferent mode deals with circumstances in which the pieces of information are weakly presented and working against each other, making it impossible to establish any plausible overarching meaning, presenting a tendency towards randomness (ibid.). Lastly, the disjunctive mode differs from the indifferent mode in that the elements of information are strongly presented, accentuating their stark opposition and therefore escalating the ambiguity

of the whole. This last mode is characterised by a tendency towards ambivalence and can evoke strong affective reactions and easily lead to interpretive conflicts (ibid.: 100).

Beyond the examination of the structural dimensions, analogous to each of these modes there would be a matching mode of reception. Therefore, in the conjunctive mode of reception the viewer would try to develop an overarching meaning; in the associative mode of reception the individual’s endless attempts to search for meaning and connections would not (necessarily) aim at producing a definite result; the indifferent mode of reception would be characterised by taking an indifferent stance, in which the openness and interplay of competing information would provide aesthetic pleasure; and, finally, the disjunctive mode of reception, best characterised as an ambivalent attitude, experiencing and reflecting on “the emotional tensions that are released by contradictory information” (ibid.: 102). Whether one opts for any of these reading modes, I hypothesise, will have to do partially with a series of personality traits and characteristics, some of which will be explored in depth further below.

One interesting aspect of Krieger’s model is the understanding of aesthetic ambiguity as a multidimensional phenomenon, not to be taken as definite categories but, rather, more like poles in spectra. The model offers a terminology that can prove useful in describing ambiguous artworks, especially in analyzing the similarities and differences among various texts. Moreover, the consideration of both the dimension of production and reception of ambiguity is key to understand the films under study in these pages.

Another interesting relation emerges in examining the phenomenon of ambiguity, not between the production and the reception, but between the cultural climate and the production. Shifting slightly the focus on the analysis of ambiguity’s presence in the arts, Berndt and Kopenick suggest an understanding of ambiguity as response. On this wise, they argue that the escalating complexification of our world, on the social, economical, political and cultural fields, might explain the need for new aesthetic strategies of presentation and representation (Berndt & Koepnick 2018: 7). These strategies would serve to illustrate the inefficiency of the measures relying on normative expectations for disambiguation to address “the complex entanglements of our contemporary world” (ibid.: 8). Following this, the increasing presence of ambiguity in post-classical narratives could be seen as part of said aesthetic reaction to our sociocultural moment, characterised by an ever-growing distrust in media representations.

Going back to Krieger, her overview of the conceptualisation of aesthetic ambiguity also identifies how several theorists have established this connection between the notion of

ambiguity in art and the historical context in which it arose, interpreting it as “a form of processing or reacting to social and cultural circumstances” (Krieger 2018: 80). Furthermore, Krieger stresses the meaningful contribution of Umberto Eco in first laying the foundation for historicising the phenomena of artistic ambiguity, while at the same time connecting it to systematic reflections (ibid.: 81). Understanding ambiguity as the result of the breach of established conventions of expression, Eco further suggests that, from a cultural point of view, the formal dimensions of artworks could be deemed as epistemological metaphors. They function as such in that they exhibit analogue structures to the Zeitgeist in which they were created, the repercussion or representation of a “widespread theoretical consciousness” (Eco 1962: 87).

With regard to some of the defining characteristics of our current cultural/social climate, May Ien Ang argues that the idea of the world being equivocal and complex is “now an important part of the Zeitgeist, and emergent, twenty-first century structure of feeling” (Ang 2011: 781). A system heavily characterised by unclear interrelations, inconsistency in the relationship between different elements and the lack of unambiguous meaning and predictable, linear relations of cause and effect” (ibid.: 782); a consequence, she argues, of the accelerating globalisation we are part of. In this context, Ang differentiates between the difficulties and the messes populating our world: the former implying the agreement about the nature of the problem, while the latter entails a lack of agreement on the problem’s origin, therefore hindering the obtaining of a workable solution and, as a result, leading to high levels of uncertainty. A key aspect of a messy problem is the multiplicity of valid perspectives and interpretations on the issue, a circumstance in which “the ambiguity of the situation is paramount” (ibid.: 786). It follows, then, that our experiencing of our contemporary world as greatly complex would have to do with the “proliferation of escalating messes” (ibid.: 787), creating an environment in which a lot of the problems we face are short of a straightforward solution, oftentimes requiring people to take ambivalent stances in order to tackle certain issues in an as effective as possible manner.

Beyond this complex outlook on our (complex) contemporary world, there is another emerging peculiarity of our Zeitgeist that can be seen to influence this dialogue imbuing the arts (in particular filmmaking) with ambiguity. Namely, the growing unreliability on media communication, which is facing a rather uncertain future in light of new technological developments in the recent and upcoming years. One that is drawing a lot of attention recently

are the so-called deepfakes, digital manipulations of audio or video reaching such a level of 1 realism that its detection of authenticity is proving increasingly difficult. They have appeared as a result of advances in the artificial intelligence known as deep learning, that is, systems in which algorithms (neural networks) learn to infer rules and eventually are able to replicate patterns by analysing large data sets (Chesney & Citron 2018: 148). Deepfakes signify a considerable upgrading from previous deceiving practices such as shallowfakes, which are essentially videos made without the use of sophisticated AI, but mainly relabelling and recontextualising them to support specific claims, sometimes presenting simple edits in audio tracks (Gregory 2019: n.p.). In the case of deepfakes, the AI works by pairing algorithms against each other in GANS (generative adversarial networks), which contains a “generator” algorithm and a “discriminator” one: the former creates the content modelled on source from large data sets, while the latter aims at spotting the artificial content, a disposition which leads to rapid improvements in the final results (ibid.). Invented by graduate student Ian Goodfellow in 2014 as a procedure to generate new data from existing data sets, this machine learning technique can be used to create fake videos (or audio or text) as long as it has access to a broad data base containing references of the content to be replicated (Schwartz 2018: n.p.). Similar to these deepfakes videos, Google presented in 2018 their AI voice chatbot technology called “Google Duplex,” which can generate natural-sounding speech reaching such a level of realism that they have been tested to, for example, successfully schedule appointments on their own and passing seemingly undetected by their interlocutors (Bates 2018: n.p.). One particularly distressing concern about these developments is that they exploit our biological predisposition, that is, the fact that we as humans are hardwired to trust our eyes and ears, because nothing persuades “quite like an audio or video recording of an event” (Chesney & Citron 2018: 147) as it provides people with the ability to become firsthand witnesses, avoiding the slippery process of assessing the reliability of someone else’s narration. This inherently natural tendency to believe what we see, explains Christoph Bietz, is what designated television originally as the “Fenster zur Welt — the window to the world” (Bietz 2015: 293), however, the unreliability of the fake-immediacy of this window (and its numerous digital duplicates) is becoming more prominent. Along the lines of this biological predisposition, film scholar Torben Grodal argued against the common tenet of ‘suspension of disbelief’ in film experience, contending that fiction actually demands a

Term emerging from the combination of “deep learning” and “fake videos”.

“modification of belief or a suspension of belief so that film viewing does not produce full-scale illusions” (Grodal 2009: 154). Therefore, the viewer would not need to make an effort in order to believe the events depicted by the moving images, but, rather, to not believe in them, to be able to distinguish it from reality itself due to, Grodal argues, our default mode of “seeing is believing” (ibid.).

There are various notably worthy applications that this deepfakes technology could be put to, such as education purposes or speech restoration, however, as law scholars Chesney and Citron point out, a series of darker purposes will (some already have) enter the game, such as digitally altered pornographic videos, blackmail and intimidation, and, probably the most troubling on a global scale, its entry into the realms of politics and international affairs (ibid.: 148). Without disregarding the positive applications of deepfakes, they are 2

(alarmingly) likely to contribute to the further deterioration of the current epistemological decay, inasmuch as they arrive in an already delicate Zeitgeist where the line between fact and fiction is becoming increasingly elusive, and where the media environment “interact in toxic ways with our cognitive biases” (Schwartz 2018: n.p.). Along these lines, Sam Gregory, technologist, media-maker and advocate of visual and participatory technologies for human rights, draws attention to the fact that the responses to the phenomenon of deepfakes need to be considered in relation to already existing misinformation trends, “from bots and computational propaganda to the online attention economy to declining public trust in institutional information” (Gregory 2019: n.p.). It is easy to imagine how these media tendencies will worsen with the appearance of highly refined audiovisual manipulations, an aggravation that in turn might lead to the very bleak point of “disbelief by default” (ibid.), that is, the moment in which the society will accept the unreliability of audiovisual media information by default. It is at this point, Gregory claims, where deepkfakes could turn into a weaponised technology, of great aid for dogmatic governments around the globe (ibid.). In a like manner, Chesney and Citron predict that the progressive refinement of the technology will result in an increment of “false yet highly realistic audio and video content, ready to be weaponized” (Chesney & Citron 2018: 155), for instance, in political campaigns and social media, the most immediate consequence of which will be the hurting of the democratic fabric.

On his article for The Guardian, Oscar Schwartz explains how “the use of this machine learning technique was

2

mostly limited to the AI research community until late 2017, when a Reddit user who went by the moniker ‘Deepfakes’ […] started posting digitally altered pornographic videos. He was building GANs using TensorFlow, Google’s free open source machine learning software, to superimpose celebrities’ faces on the bodies of women in pornographic movies.” See (Schwartz 2018)

Citron proposes the hypothetical scenario in which a deepfake is made public the night before an election date: while technologies are being developed that could eventually trace the inauthenticity of the content, most likely the damage will have already been done (Schwartz 2018: n.p.). Similarly, not even the often valuable contributions of citizen journalism will be exempt from the effects, due to the spreading of the “liar’s dividend,” that is, the more aware people become of the existence of deepfakes, the less likely they are to believe in any audiovisual piece, which in turn will make it easier for the protagonists of the piece to cast a doubt over its legitimacy (Chesney & Citron 2018: 151-152).

Hany Faris, Professor of Computer Sciences at the University of California, believes that researchers should be more cognisant of the social and political ramifications of their inventions, examining the balance of positive and negative implications (Schwartz 2018: n.p.). Concentrating on the development of forensic technology to deal with deepfakes, Faris claims that these type of detection mechanisms need to work in tandem with a deeper understanding of how this phenomenon alters the way people processes information. The plausible consequential growth in distrust towards audiovisual content would entail a threat towards the “processes by which we collectively come to know things and hold them to be true or untrue,” in which not the falsity of the content per se, but the mere prospect of their existence is the most harmful, what scholar Aviv Ovadya calls “reality apathy” (ibid.). Therefore, it is essential to clearly understand the way people receive and share information, because, as researcher Hwang explains, it is primarily by gaining a better understanding of the behavioural factors driving us to believe in hoaxes that we stand a chance to avoid “an unwinnable game of technological whack-a-mole” (Hwang 2018: n.p.). Correspondingly, Gregory underscores — in addition to a series of technological approaches — the importance of analysing our patterns of communication: understanding people’s engagement with (and sharing of) misinformation, recognising our cognitive limitations and predispositions, as well as acknowledging the real influence of objectivity as a criterion for trusting and sharing content (Gregory 2019: n.p.). All this intended to aid in developing the new literacies the public will be needing in order to “prepare for the first wave of malicious synthetic media” (ibid.).

The majority of the theorists writing about deepfakes advocate for a proactive attitude that moves beyond the initial discouraging shock of pessimism brought about by the uncertainty of such a future outlook. This is something they share with Ang’s approach to her theorisations about the complexification of our world discussed above. In this fashion, she

insists that complexity in itself should not be the message, or at least not the only one. While identifying the problem is crucial, more needs to be done beyond simply recognising how “complex and contradictory” reality has become (Ang 2011: 785). Developing ways to learn to cope, deal with or navigate the concrete complex realities we are confronted with, what she defines as cultural intelligence, that is, the ability to “stress the constitutive role of culture in our efforts in making sense of, and dealing with, complexity” (ibid.: 790). Cultural intelligence is grounded on the premise that this navigation can never be based on unambiguous solutions, instead, it requires the painstaking negotiation of its multiplicities, different perceptions and divergent interests (ibid.). It favours “process-oriented approaches to problem-solving,” which would be based on strategies that always keep room for contingency and uncertainty along the way, in opposition to predetermined and categorical formulas.

The complexification of our contemporary climate, as explained by Ang, and the emergence in this Zeitgeist of phenomena such as deepfakes, contribute to the proliferation of a certain obscurantism in our daily consumption of media content. The resultant constant presence of ambiguity in media information and its effects on the audience, moreover, foregrounds a growing unreliability to what we see and hear in acts of mediated communication. And it is precisely here where a strong similarity exists in relation to the trend of contemporary ambiguous films this text is concerned with. Without disregarding the obvious differences between both scenarios, the ambiguating of post-classical narratives in cinema can be perceived as a direct reaction to a very palpable reality, one where indefiniteness is becoming ever more predominant, imbuing their narrations with processes of hypothesis creation that are heavily challenged or even never resolved.

Throughout the previous pages I’ve been using the term unreliability/unreliable without really giving a proper definition of the concept, somehow placing it in a common context with ambiguity. Let us quickly nuance this before moving forward.

Among many available, Gregory Currie proposes a definition of unreliability understood in terms of the complexity intended by the implied author (Currie 1995: 23). The main advantage of this definition (in comparison with that originally presented by Booth) is that it allows to assign it to a narrative as a whole, even when there is not a clear source of unreliability within the diegesis in question. What is more, Currie examines the relation between unreliability and ambiguity, concluding that while they are distinct concepts, they are highly compatible: unreliability does not need ambiguity, but “it is an easier effect to achieve

when it goes with ambiguity than when it does not” (ibid.: 25). The main challenge for the unreliable narration is to set a series of clues at two (or more) different levels (ibid.). In the ambiguous narrations of some of the case studies, the different contradictory levels confront each other directly, not simply during the later process of solving the “epistemic distance” (ibid), but, rather, during the course of the film.

Relevant for our analysis — and complementing the previous varied definitions of ambiguity — Currie posits a criterion to determine when a narration is ambiguous, namely, “when it raises a question in the viewer’s mind which it fails to answer, and where the raising and the nonanswering seem to have been intentional” (ibid.: 24). In addition to this, another aspect of these narrations should be considered, that of the degree of contradiction in the answers offered to these questions. So, alluding to Zeki’s “certainty of multiplicity” (Zeki 2004: 188), we could understand an ambiguous narration as that which presents a series of (1) either intentionally unanswered questions or (2) consistently answered in (occasionally) mutually contradictory ways. This approach to the narrating of the story arises in the viewer a sense of unreliability, which sometimes might be possible to assign to diegetic narrators, but which more generally permeates the film narration (artefact) as a whole. This unreliability of the narration may inevitably point towards the presence of an “editorial intelligence” (Tan 1996: 65), what Currie would consider the unreliable implied author. And, in acknowledging this operating intelligence, we would consequently become aware of the film as an artefact, carrying with it a series of particular “artefact emotions” (ibid.), that is, reactions directed towards the filmic device in itself, rather than to the events of the narrative.

3. Individual Differences and the Arts

3.1 Personality Traits

Meanwhile, these films can be seen to provide to some extent practices of familiarisation with if not coping strategies against instances of ambiguity, a sort of cognitive playground within the safety of the arts. This cognitive playground would be concerned with putting to work a 3

series of personality traits fluctuating among individuals, and which affect to varying degrees

Kiss & Willemsen argue that the contemporary trend of cinematic complexity could be understood as a

3

“narrative instrument aiming to provide a cognitive playground” (Kiss & Willemsen 2017: 27). The films considered here could be seen to form a narrower sub-trend within the more overarching trend of contemporary complex cinema studied by Kiss & Willemsen.

the aesthetic experience of films. Some of these traits play a key role in both diegetic and extradiegetic contexts when it comes to facing ambiguity and dealing with unreliable narrations. Although a proper research in this direction will require the collaboration with media psychologists, a brief summary of studies concerning psychology attributes such as negative capability, need for cognitive closure, or (in)tolerance of ambiguity/uncertainty will enrich the present analysis of contemporary ambiguous films (and their reception).

Psychologists have shown interest in examining the relation existing between certain personality traits and arts’ preferences, as well as the capacity of art consumption in refining or training some of these skills. A significant contribution in this field was actually done by poet John Keats in the nineteenth century, in a letter to his brothers. Here, narrating a discussion he had that evening, he coined the term negative capability to refer to the capacity of a man of “being in uncertainties, Mysteries, doubts, without any irritable reaching after fact and reason” (Scott 2005: 60). This term was rescued a century later by theorist Reuven Tsur to establish a spectrum defining people’s approach to literary criticism, placing negative capability in one pole, representing high levels of tolerance of ambiguity, and quest for certitude on the opposite pole, characterised by a “psychological atmosphere of security offered by facts, reasons, labels and theories” (Tsur 1975: 776). People falling in the former group present the ability to cope with delayed closure, being capable of remaining in uncertainty, which allows them to appreciate complex/ambiguous meanings in their readings; in contrast to those presenting rapid closure tendency, inclined to provide authoritative explanations rather than making delicate distinctions (ibid.: 779). Furthermore, he compared the duality of negative capability vs quest for certitude with psychological tendencies like flexibility/rigidity, abstract/concrete personality and tolerance/intolerance of ambiguity. (ibid.: 787).

The term intolerance of ambiguity (IA) was first described by psychologist Else Frenkel-Brunswik as an emotional and perceptual personality variable, basic in the emotional and cognitive orientation of a person toward life (Brunswik 1949: 113). Frenkel-Brunswik compares the stimulus individuals receive as playing the role of an authority, therefore if this authority (stimulus) is perceived as ambiguous, it will be as “disturbing to the prejudiced as would be a leader lacking in absolute determination” (ibid.: 128). Moreover, she points to the inner confusions and conflicts in the prejudiced individual as the cause of his or her tendency to black-white solutions, which is an “oversimplified and thus reality-inadequate approach” to our world (ibid.: 134). As a result of her tests, she reveals a positive correlation

between intolerance of ambiguity and prejudice, as prejudiced individuals presented the biggest tendencies to premature reduction of ambiguous cognitive patterns. An affinity (between hostility, power orientation and rigid stereotyping and intolerance of ambiguity) that she ultimately defines as a struggle basic to our civilization (ibid.: 141).

In 1993, David L. McLain revised some of the different perspectives on the issue of tolerance for ambiguity since Frenkel-Brunswik’s contribution and defined an ambiguous stimulus as something that can be perceived as new and unfamiliar, too complex to understand or having multiple and incompatible interpretations (McLain 1993: 184). Following this triple characterization of ambiguity, he refines tolerance for ambiguity as a spectrum, ranging “from rejection to attraction, of reactions to stimuli perceived as unfamiliar, complex, dynamically uncertain, or subject to multiple conflicting interpretations” (ibid.). Among others, the tolerance for ambiguity has been found to be interrelated with issues of sociopolitical ideology (general conservatism) or people’s tendencies to seek supportive rather than objective information (Furnham & Ribchester 1995: 183-84). Furthermore, three different reactions have been distinguished in relation to manifestations of intolerance of ambiguity: cognitive reactions, involving the tendency to adopt rigid responses; emotional reactions, referring to expressions of uneasiness, discomfort, anger or anxiety; and behavioural reactions, the rejection or avoidance of such situations (Grenier et al. 2005: 594). Another similarly related concept, interesting in examining ambiguous films, is that of need for cognitive closure (NFCC), defined by Webster and Kruglanski as a person’s motivation concerning the processing and assessment of information (Webster & Kruglanski 1994: 1049). Once again, a spectrum is drawn to contemplate the opposite predisposition, that is, the need to avoid closure, which they argue could be motivated by “the perceived costs of possessing closure and the perceived benefits of lacking closure” (ibid.). It should be noted, however, that both need for cognitive closure and need to avoid closure would ultimately strive for cognitive closure, the difference being their approach to information, while the former does it with closed minded attitude, the more open-mindedly approach of the latter affords him or her to relish the process’ possible inconsistencies (ibid.: 1056). The need for closure motivation was found to affect people’s behaviours in both social cognition and interactions, in things such as the use of stereotypes, correspondence bias, readiness to be persuaded, rejection to opinion deviates and effects in interpreting ambiguous information (ibid.: 1061). Thus, it can be deduced that the position a viewer hold within this spectrum will affect the way he or she experiences the multiplicity and insolubility of ambiguous narrations.

If the viewer presents a high urge to obtain cognitive closure, he or she is likely to either adhere to a possible solution without much deliberation or maybe give up all together as a result of the frustration such delay in cognitive closure may entail.

Van Hiel and Mervielde later on examined the need for closure — the need to have an answer vs enduring ambiguity —, which is positively related with a preference for order and predictability, and induces discomfort in ambiguous instances (Van Hiel & Mervielde 2003: 559). Very significant, as Van Hiel and Mervielde argue, is the lack of interest in new information and evidence this personal disposition implies, meaning less considerations about alternative hypotheses and less extensive search for information, often leading to the use of simpler cognitive structures (ibid.: 560). That is, individuals with a high disposition of need for closure are inclined to dedicate less effort to the weighing of contrasting alternatives in equivocal situations. This manner of processing information is the result of a combination of two tendencies, urgency tendency and permanence tendency, referring to the desire to attain closure quickly and the desire to preserve previous knowledge, respectively (Webster & Kruglanski 2011: 139). Webster and Kruglanski argued that this predisposition presents a series of social-cognitive effects: less extensive information processing and hypothesis generation, elevated judgmental confidence, preference for prototypical structures or greater reliance on pre-existing knowledge structures; as well as interpersonal consequences: lowered emphatic concern for dissimilar others, increased rejection of opinion deviates and greater tendency to task-oriented vs. socio-emotional attitudes (Webster & Kruglanski 2011: 162-163). The ambiguity in film narratives create situations that have a high potential for “adaptive flexibility” (Van Hiel & Mervielde 2003: 566), where the processes of inferencing and hypothesis making are met with obstacles, prompting the viewers to change their strategies. In such instances, the need for closure carries out a counterproductive role inasmuch as it would motivate the viewer to achieve a rapid solution (urgency tendency), detracting from the whole experience of weighing alternative possibilities (permanence tendency).

Another concept usually considered in combination with the already described is the intolerance of uncertainty (IU). In a similar manner to intolerance of ambiguity, intolerance of uncertainty refers to the individual’s negative assessment/predisposition to ambiguous situations and his or her tendency to avoid them as sources of stress and upsetting emotions (Grenier et al. 2005: 595). Consequently, groups of people presenting high levels of uncertainty intolerance will tend to be more ritualized and task oriented (Furnham &

Ribchester 1995: 194). The fundamental difference between these two concepts is that of time orientation, particularly, intolerance of uncertainty refers to the unpredictability of an event that is forthcoming, whereas intolerance of ambiguity is concerned with ambiguous stimulus in the “here and now” (Grenier et al. 2005: 596). The source of threat and discomfort is either placed on the present (for IA) or in the future (for IU): the avoidance of ambiguity, then, would seem to precede uncertainty. Within the context of aesthetic experience, high levels of uncertainty avoidance could be expected to affect the individual’s tendency to rigidly adopt certain genre-specific viewing schemata when watching a film, considering how this can act as a guide to partially predict the unfolding events taking place in the narrative.

3.2 Aesthetic Empirical Research

Following this quick conceptual examination, let us now look at some of the recent research conducted relating art and aesthetic preferences with diverse personality traits. In 2009, Chamorro-Premuzic and his team carried out a study aiming at establishing connections between subjects’ preferences in painting artistic movements and individual differences in personality (Chamorro-Premuzic et al. 2009). From the Big Five personality factors, openness — which measures intellectual curiosity, creative interests, preference for new experiences and “toleration of the unfamiliar,” and is generally distinctive of individuals with high levels of imagination, need for cognition and divergent thinking (ibid.: 502) — was the strongest and only consistent personality predictor of artistic preferences according to their results (ibid.: 511). This adds to previous research positing that open individuals display more 4

positive aesthetic attitudes towards non-conventional art, without necessarily achieving a complete emotional understanding, that is, showing an appreciation for intrinsic rather than literal representations (ibid.: 503). In addition to this, they found that age is another decisive factor. In this way, they argued that art preferences are “only partly style-specific,” being possible to some extent to predict aesthetic inclinations based on psychological disposition and age (ibid.: 513).

Moving to the field of written narrative, a study was conducted in 2013 examining the relation between reading fictional short stories (in contrast to nonfictional essays) and the possible reduction of need for cognitive closure this might cause (Djikic et al. 2013), tapping

The “Big Five” is the encompassing term employed to refer to the main personality factors. These are:

4

into the idea of arts as cognitive playgrounds, particularly the potential of fiction narrative to opening minds. The main hypothesis was supported, i.e., that participants experienced a significant short-term decrease in need for cognitive closure when reading a literary short story, when compared to reading an essay (ibid.: 153). They argued that one reason for this opening the mind responds to the type of thinking a person engages with while reading fiction: on the one hand, the reader’s concerns regarding urgency and permanence are discharged, while still offering situations closely related to real life through somebody else’s prism. Their results showed that this double release — thinking without concerns of urgency and permanence and thinking in ways different to one’s own — occurred mainly by a decrease in need for order and a decreased discomfort with ambiguity (ibid.). Moreover, they suggested that one cause of the decrease in need for closure might be the meta-cognitive processes involved in thinking from a different perspective (ibid.), which would explain its difficulty of opening the mind. However, it remains unclear whether this reducing effect extends beyond the fictional scope, making it an investigating path worth pursuing considering the potential benefits for society.

Focusing now on the field of moving pictures, professor of social psychology Viren Swami and his team set out to further explore the relation between personality traits and the preference for more complex and challenging aesthetic compositions in film (Swami et al. 2010: 856). For that, they evaluated the Big Five personality factors, sensation seeking (desire to seek out varied, complex, novel and intense experiences) and tolerance of ambiguity while watching surrealist films — considering them ideal based on their “incongruous and ambiguous imagery and situations” (ibid.: 855) — which viewers had to rate according to their personal preference and familiarity. As a result, openness, sensation seeking and tolerance of ambiguity were significantly associated with a preference for surrealist motion films (ibid.: 858). Additionally, they put forward the practical benefits of these results, in terms of promoting art in the community, its use for therapy and as a theoretical proof that these personality traits have real world consequences (ibid.: 859).

Along similar lines, film scholar Kovács and psychology scholar Papp-Zipernovszky very recently explored the mental strategies a viewer is confronted with when the causal inference process is challenged in the reception of a narrative film and its potential relations to personality factors. They consider Bordwell’s cognitive theory of film in order to conduct the experiment, therefore a brief detour for a rapid recapitulation of the constructivist account

presented in Bordwell’s Narration in the Fiction Film (1985) is in order, especially as it will be used later on when examining the ambiguity of the case studies.

Bordwell’s understanding of film viewing presents it as a dynamic process, managing a variety of factors: perceptual capacities of the viewer, prior knowledge and experience (prompting the use of schemata) and the structural dimension of the film itself; taking the construction of an intelligible story as the main cognitive goal (Bordwell 1985: 32-33). As a main guiding tool, viewers presuppose a master schema, which represents typical expectations about classification of events and the way the individual parts relate to the whole and the canonical story format, which is the most common template structure (introduction; explanation of state of affairs; complicating action; ensuing events; outcome; ending) (ibid.: 35). According to Bordwell, viewers would have a series of expectations regarding the narration: an expository fragment at the beginning, the complication of a state of affairs, and following a character functioning as a goal-oriented protagonist. Based on this, the viewer justifies the filmic material following several possible motivations, two of them are more common: compositional (relevant to the story) and transtextual (genre tropes), with two less common ones, the realistic (based on knowledge about the world) and the more rarely artistic one (for the sake of it) (ibid.: 36).

In this context, Bordwell argues that audiences perform a series of cognitive activities during the process of figuring out the narrative: first, the viewer presents a series of assumptions (be it about our real world knowledge or particular to a specific genre); second, the viewer also makes inferences about the events on screen; and finally, the cognitive task of hypothesising, where “the spectator frames and tests expectations about upcoming story information” (ibid.: 37). This process of hypotheses making is prompted in different ways depending on the structure of the narration. Bordwell characterises hypotheses in a set of opposites: curiosity or suspense, the former refers to past action the narration does not specify, while the latter sets up anticipations about future events; more or less probable, moving from the highly likely to the flatly improbable; more or less exclusive, ranging from either/or choices to mixed sets of them; and either successive, when one hypothesis is being replaced by another, or simultaneous when several hypotheses are kept at the same time (ibid.). In addition, these hypotheses can be tested in a microscopic or as a macroexpectation level, the former signalling the moment-by-moment processing, while the latter encompasses a wider arc (ibid.). In instances where the narrative blocks hypothesis validation (or invalidation),

Bordwell argues that the viewer adopts a “wait-and-see” strategy, hoping that the forthcoming events will aid in this.

Important for this constructivist account of narration in film is the triad consisting of fabula, syuzhet and style. The fabula is the pattern subjectively constructed by the viewers as a result of the above-mentioned cognitive operations, and on the basis of varied schemata: prototype schemata (types of persons, places, actions, etc.), template schemata (the canonic story) and procedural schemata (a predisposition in the search of motivations and relations of causality, time and space) (ibid.: 49). The syuzhet (commonly known as plot), is the representation of the fabula of the film, providing a “set of cues prompting us to infer and assemble story information,” guiding us in the construction of the said fabula, eliciting curiosity, suspense and surprises (ibid.: 52). And lastly, the style refers to the cinematic devices employed by the film in a somewhat systematic manner (ibid.: 50). All these three systems combine in the construction of the narration, that is: “the process whereby the film’s syuzhet and style interact in the course of cueing and channeling the spectator’s construction of the fabula” (ibid.: 53).

This understanding of the narration, differentiating between fabula, syuzhet and style, facilitates analyses of film narratives particularly in instances when the narrations turn ambiguous, and assists in comprehending the stimulation of certain cognitive activities in the viewer, presenting cues and key information contributing to processes of causal thinking, inference making and hypotheses testing. One of the ways in which this is done, is via the careful selection and combination of the fabula events by the syuzhet, which in turn creates a series of gaps. These are among the clearest cues for the viewer according to Bordwell, evoking the entire process of schema formation and hypothesis testing (ibid.: 54). These gaps can be: temporary or permanent, depending on whether the information is provided or not; diffuse or focussed, whether they lead to non-exclusive or exclusive hypotheses; and flaunted or suppressed, whether the syuzhet calls attention to the gap in question or not (ibid.: 55).

Having briefly considered some relevant aspects of the cognitive theory of film, let us return to Kovács and Papp-Zipernovszky’s recent experiment, which focuses on the process of causal inferencing in narrative film. A first distinction is established between two processes comprising causal relations in real life, namely, causal perception and causal thinking, being the latter the focus of the study, as they consider that the link between personality traits and the causal understanding of film is likely to happen (mainly) at a “high-level causal thinking part” (Kovács & Papp-Zipernovszky 2019: 6). With this in mind, they examined what

happens when the causal cues are missing or inconsistent in a film narrative and whether mental approaches different from the “wait-and-see” strategy suggested by Bordwell enter the game, particularly in narratives lacking unambiguous time-space-casual cues (ibid.). Discerning between classical (character-centered, goal-oriented and with causality-driven narratives) and nonclassical narratives (causality is not the main principle and character’s goals are unclear), they screened A Man Without A Past by Finnish director Aki Kaurismäki (2002) — as an example of the nonclassical mode in a first study — and two short films — each corresponding to each of the narrative modes — in a second study, monitoring the participants’ inferential activity during the viewing experience.

The results suggest that the viewer’s causal construction is only partly predictable by the text, with the personality traits of Openness to Experience and Meaningfulness of Life proving determining in the amount of causal inferences made by the viewer; the effects of which are stronger in “dramaturgically less important moments” (ibid.: 18), that is, in moments where the syuzhet is not cueing so strongly cognitive efforts on the audience. Similarly, the more narratively incoherent the film is as a whole, the more decisive they found these traits to be (ibid.: 26). The registering of a high matching number of viewer’s causal responses during key sequences of the film confirms Bordwell’s hypothesis of the “active viewer,” proving that indeed the viewers react to the cues provided by the narrative (ibid.). Nevertheless, and contrary to their expectations, the causal responses were significantly higher during the viewing of classical narratives as opposed to the nonclassical ones, where they predicted an equal or higher amount of mental causal activity (ibid.: 27). In light of this, Kovács and Papp-Zipernovszky argue that viewers refrain from causal inferences once they realise the unconventional nature of the narrative structure of this type of films (in the nonclassical examples) (ibid.). This can be a result of the gradual reduction of understanding resulting from the succession of “ambiguous causal connections” (Bordwell 1985: 35), eventually inviting the viewer to stop altogether.

Regarding the classical narratives, and despite its strong causal structure, in many points throughout the narration they observed considerable differences in the amount of causal activity, situations in which the causality was considered to be largely independent of the text’s structure and more dependent on personality traits, which could account, they argue, for interpretative differences (Kovács & Papp-Zipernovszky 2019: 27). As a concluding remark to their study, Kovács and Papp-Zipernovszky suggest that in nonclassical narratives the viewer’s personality perform a stronger contribution, and that these films do not seem to

necessarily demand more mental activity but, rather, a different type, a cognitive processing that might not rely so heavily on causal connections (ibid.: 28). This opens up fascinating paths for future research, in which, in spite of the “heavily inference-dependent nature of human behavior” (Tooby & Cosmides 2001: 19), different mental strategies seem to be adopted as response to the equivocal, unreliable and deceptive nature of these representations.

Once we have looked into some of the main ideas of Bordwell’s cognitive theory, it seems only natural to contemplate Bordwell’s understanding of toleration of ambiguity. He defines it as one of the strategies developed by viewers to deal with deviations from the master schema (Bordwell 1985: 35). To this extent, certain structural features of the narration will serve as referential points from which to assess and deal with ambiguous narratives. The toleration of ambiguity here is then regarded as a sort of refinement mechanism of the master schema in order to fit the narration communicated by the obscure syuzhet.

During the close reading of the case studies more dimensions of this cognitive theory of film will emerge, but for now another brief addition will be pointed out, namely Edward Branigan’s contribution in his Narrative Comprehension and Film (1992). Branigan proposed a hierarchical classification of levels and agents of narration, and in this context posited the notion of focalisation in analysing the way specific information is provided in the narration. Focalisation does not involve a character necessarily speaking or acting but, rather, “actually experiencing something through seeing or hearing it” (Branigan 1992: 101), extending to more complex aspects of experience, such as thinking, remembering, interpreting, fearing, believing or understanding (ibid). Two main types are to be distinguished here: external focalisation and internal focalisation. The former comprehends instances representing the character’s awareness but “from outside the character” (ibid.: 103), that is, we see/hear the same but not from the same spatial position, sharing their attention, not their experience. Internal focalisation, on the other hand, corresponds to the private and subjective experience of the character, stretching from simple perception (POV) to deeper thoughts (hallucinations, dreams or memories) (ibid.). Branigan’s definition of narration, then, comes to the overall regulation and distribution of knowledge, which determines when and how a reader acquires knowledge from a text. Furthermore, Buckland and Elsaesser, from Branigan’s theory, constructed a typology of four types of shots (Buckland & Elsaesser 2002), consisting on: objective shots, a shot motivated by an agent outside the film’s diegesis; externally focalised shots, a shot focalised around a character’s awareness of the events, such as over-the-shoulder shots; internally focalised shots, representing the character’s visual experience, such as POV

shots; and internally focalised shots, representing the character’s internal events, dreams, hallucinations or memories (ibid.: 190). This typology will be useful to identify specific shots in the films discussed hereafter.

At this point, a deeper examination on some of the theoretical discussions concerning the notion of ambiguity has been presented, as well as its potential commonalities with unreliable narration in general. To this, followed a brief sketch of some of the psychological terms and personality traits related to ambiguity and uncertainty, looking at some of the studies conducted connecting these traits to aesthetic experience — interspersed with the basics of Bordwell and Branigan’s cognitive theory of film. In the following I will test out these theories on the examination of the case studies, providing detailed analysis of certain excerpts to illustrate the different strategies presented characterising these contemporary ambiguous films.

4. Examining Contemporary Ambiguous Films

4.1 Open Endings

First of all, a clear distinction should be made in order to differentiate the ambiguity created as a somewhat persistent experience throughout the entire film by a series of formal strategies (each of which I will be focusing on individually), and the more commonly considered ambiguity of open endings. In discussing narrative closure, Noël Carroll talks about erotetic narration, a narration that follows a path built upon the logic of positing questions that are subsequently answered (Carroll 2009: 211). The narrative closure would be obtained when all the questions that the narration presented before us have been answered (ibid.). Regarding the questions proposed by the narration, Carroll discerns (presiding) macroquestions from microquestions, the former being the dominating, main concern of the narrative throughout, while the latter are of a more localised type, with a more limited scope (ibid.: 212). Therefore, the narrative closure would be attained when, Carroll argues, the macroquestions are answered by the narration. While violating the erotetic narration can render more lifelike, providing “a rhythm more akin to the everyday flow of often directionless events” (ibid.: 214), most pictures intend closure, often especially because of this same reason, to contrast our more “desultory and diffuse mundane lives” (ibid.: 215), as well as, Carroll observes,