Narrative Voice/over in the Essay Film

James V. Catano

Louisiana State University

“How and why did documentary narration acquire its miserable reputation whilst still remaining one of the most commonly used devices in nonfiction filmmaking?” (Bruzzi 2000: 47)

“[T]he essay film [is] a practice that renegotiates assumptions about documentary objectivity, narrative epistemology, and authorial expressivity within the determining context of the unstable heterogeneity of time and place.” (Corrigan 2011: 6)

“[T]he imaginative use of voice-over and voice-on narration […] enables [film] to problematize the relation of discourse to story, of narrator to narrated, of imagination to reality.” (Chatman 1999: 337)

The history of narration and documentary film, increasingly known as nonfiction film, is not an altogether happy one. But that is gradually changing, with ongoing and particularly interesting narrative experimentation happening in a subset of nonfiction prose and film: the essay. With its roots in Montaigne’s writing, the ‘personal essay’, as it is often called, provides a particularly fertile ground for discussing both the concerns and actions of nonfiction and their relation to narrative play and narratology. Together, these three areas – essay, film, narration – constitute the basic concerns of this discussion. Unified by conceptual forces and cultural desires, this constellation of issues pulls further related terms and concerns within its overall orbit: persona, voice, and point-of view or focalization among them. Together, the terms provide multiple entries into “essaying,” in the Montaignean sense, the nonfiction film essay and the particular role it plays in narrative thought and practice.

I want to initiate the discussion by providing historical context for documentary as well as film essay practice, because these practices already operate within a varied set of assumptions surrounding the ‘nature’ of nonfiction narrative. I also want to provide background on the prose essay and suggest that, while it seems to be a form that ostensibly lacks a story per se, the essay nevertheless has what can be considered a narrative progression within it. Finally, I intend to bring together these two areas, narration and non-fiction film, in the form of the contemporary film essay. To do so, I will take a look at the films of two individuals, Agnès Varda and Ross McElwee, as a way of demonstrating how the issues of narration and essay come together in interesting forms in the medium of film.

1. Narrative and nonfiction

I start with one basic given: oral narration is central to much documentary. Because such narration may occur in the form of voiceover or voice-off, non-fiction film often relies upon extensively scripted narration. In many documentaries spoken narration may even serve as the driving force of the film, with the visual materials there primarily to illustrate what is orally narrated. Indeed, Bruzzi’s description of narration’s “miserable reputation” in nonfiction film may very well result from it’s being “one of the most commonly used devices.” If popular expectation of cinematic experience is one of being enveloped in visual stories, then the speaking presence of a narrator may well be seen as intrusive. ^[1]

The history behind these overall attitudes will be addressed below. But recognizing the presence of an extra-diegetic narrator as a central component of much nonfiction film – traditional and essay – makes it easier to move forward and take up narration’s role. At the same time, while most documentary film triggers a sense of a “teller,” it may not look like the “story-teller” of Kellogg and Scholes’s classic definition of literary narrative. Such differences among genre are inescapable, as attested by attempts in narratology to deal with cinematic and written fiction. But the narrator of the essay film does appear to more directly overlap with the concept of character-oriented narration in ways that open up to narratological discussion. The goals here are not to solve these cross-media issues, but rather to see how they might be addressed, in this case by looking at a genre not overly theorized yet, the film essay.

To start, it is worth noting that the teller of literary narrative has two near-relatives in the prose essay, that of character and persona. In fact, persona can become, as regularly happens within literary narrative, a character in its own right. Seen in this light, the narrative provided in the essay may in no small part be seen as the story of narrating. Said another way, the narrator of the essay exists in the text as a character whom the viewer watches and hears deliberating on the subjects at hand, among which are statements on the activity and implications of creating the text itself. In the essay, and particularly in the film essay, the act of making the essay is the ‘story’ of the narrative, as will be argued below. But for now, it is enough to posit that the narrator is a central component of much nonfiction film – traditional and essay – so as to move on to take up Bruzzi’s concern with narration’s miserable reputation within nonfiction film in general, and finally and most particularly, the question of narration within the nonfiction essay film.

2. Narration and documentary film: context I

There are innumerable answers for the reputation of narration in nonfiction film. But most responses, including Bruzzi’s own, note two key trends within the history of film in general, as well as in nonfiction film – popularly known as documentary – which have given rise to this paradoxical hatred/ongoing usage of narration. The first trend, which goes all the way back to the conflict between silent films and the early days of sound cinema, rests in a belief that the visual image is itself the most authentic or genuine form of filmic representation and filmic art. It is ‘camera-narration’ that establishes the true nature of film as an art form.

Within fictional films, this argument for a pure visual narration, unmarred by the presence of dialogue, finds its high water mark in films such as La Passion Jeanne d’Arc. While the film is itself an historical recreation that draws extensively on the proceedings of Joan’s trial, it harbors no goals of being seen as documentary per se. Nor does it intend to signal its own heavy reliance upon written sources: trial documents, historical accounts, or any other form of representation. The visuals are the thing wherein they’ll capture the essence of a martyr – a feat they very successfully achieve.

Passion is, among other things, a primer on visual narration. Even with intertitles and any assumed extradiegetic musical accompaniment, the film demonstrates how to narrate a story visually. But at the moment it was appearing, arguments for a purely visual narrative were falling before the onslaught of the talkies. Sound would win out within the world of popular, primarily fictional films, with emphasis on voice-on sync-sound and dialogue in service to a representational mimesis. When they occurred, subsequent silent films would arise now and then to comment on the prevailing trend. 

The second major trend lending itself to the problematic history of voiceover or voice-off oral narration can be found in documentary’s early affiliation with so-called voice-of-god narration. Such narration was rooted in the socially driven goals of John Grierson’s early desire to align ‘documentary’ films (which he is credited with naming) with an informed citizenry. This sense of the didactic possibilities of film too readily matched with a sense of the lecture as the most viable form of pedagogy. Firmly established between World War I and II, such attitudes reached full fruition in the nationalistic goals of World War II documentaries, with their desire to garner support for a war effort from amongst a wide populace not necessarily well-informed about the logistical and ideological goals of their respective nations. The Why We Fight Series offers numerous examples:

There are many things to dislike about this narration, both in its specific statements and its overall formal structuring. Morally, the racism is repugnant. Civically, the use of blatant emotionalism to override complex thought is disturbing. Cinematically, the reduction of visual narration to mere graphic illustration of the voiceover is disheartening. Taken as whole, the clip readily demonstrates why narration in documentary film acquired a “miserable reputation” that it is still trying to shake. 

There have been, of course, a variety of challenges to that tradition. In the United States, one notable trend hints at early beliefs in the visual image as the purest form of narrative or, in the case of journalistic goals, the purest form of truth. Robert Drew’s regularly misnamed cinéma vérité, now more aptly called “direct cinema,” demonstrates this revisiting of image over word. Drew’s conceptual framework, developed with the Maysles, D.A. Pennebaker, and others, sees vocal narration as what you provide when visual narration (or the filmmaker’s mishandling of it) is inadequate to the task at hand: visually capturing reality. The narrator of the nonfiction film is thus a clumsy substitute by a director and a cinematographer who have inserted an extradiegetic oral narrator between the audience and the ‘reality’ that should unfold visually before it.

Given such an argument, direct cinema does not challenge a goal often popularly ascribed to nonfiction narrative: the representation of the real. Rather, what direct cinema challenges is the way in which that real is to be represented, that is, via a preferred visual narration rather than an extradiegetic narrator, a mode that is invested in a “non-human narrative agency” or “impersonal” cinematic narrator (Chatman 1990: 138). Direct cinema is thus not wholly antithetical to an ideology of direct and truthful representation of reality via the visual medium. To see purely is to know truly.

Cinéma vérité of the European stripe had a different sense of the medium. Influenced by language and narrative theory, as well as by healthy skepticism of the use of media for nationalistic purposes, such cinema traced its sense of truth to ideas more akin to those of Dziga Vertov’s kino-pravda, ^[2] in which the medium is never a simple translation of the material into the conceptual, but always a construction of it, in some ways superior to the human eye. Central to the rise in France of this form of cinéma vérité is the work of Godard, who directly aligned himself with such thinking and filmic production via the formation of the Dziga Vertov Group.

Importantly for the purposes of this discussion, Godard has also declared that he finds his ultimate influence in Montaigne, thus positioning his work as the cinematic realization of the goals of the ostensible ‘father’ of the prose essay. ^[3] That claim is more readily apparent in Godard’s later, less directly fictional work. Yet Godard chooses to see his entire body of cinematic production as an enactment of Montaignean goals. To see how that might be, and to establish the importance of Montaigne to narration in the film essay, it is important to establish the context provided by the history of the prose essay as a form.

3. Narration and nonfiction prose essays: context II

Over time, the term essay has become its own form of “loose, baggy monster” employed to describe a variety of nonfiction prose whose range is difficult to fully encapsulate. For the purposes of this article, essay here refers to the strain that follows directly from Montaigne and, while often possessing an ostensible focusing topic (often captured in its title: “Of Repentance,” “Of Books”), nonetheless provides a free-form, associative ‘essaying’ of the topic which regularly serves merely as a point with which to start and, less regularly, to conclude. Today, it is often characterized as the ‘personal essay’. While not unproblematic, this description at least captures the Montaignean focus on the essay’s goal of evoking, through style and voice, an active consciousness and personality to be experienced by the reader as an evolving I, a particularly fluid, decentered persona that has caught the eye of post-modernist theorists.^[4]

This decentering is the paradoxical heart of the nonfiction personal essay, paradoxical because its persona is often established as collapsing author and individual into one figure: “one may commend the work apart from the workman; not so here; he who touches the one touches the other” (Montaigne, “Of Repentance”). Yet the “workman” of Montaigne is regularly offered as fluid, a narrating subject in process of discovering itself: “What I chiefly portray is my cogitations, a shapeless subject that does not lend itself to expression in actions. It is all I can do to couch my thoughts in this airy medium of words” (Montaigne, “Of Practice”). It is a sentiment that Woolf echoes in her own reflection on the paradoxical role of the essay persona as “Never to be yourself – and yet always” (Woolf, Essays 221). Attitudes toward this multi-faceted, fluid subject have varied over time. Some authors choose to establish a full-blown narrator such as Hazlitt’s Elia, with its own directly fictive representation of elements of Lamb’s life. Others, and this matches contemporary audiences, expect the I of the personal essay to halve Woolf’s claim and ‘always be’ the author’s self. This vibration between writing self and written self is captured in an ongoing set of tensions, with the narrated/narrating self being alternately revealed and represented, authentic and artistic, innate and invented, embodied and expressed, depending as often on the desires of the audience as on the form itself.

If that is the figuration of the persona in the essay, then its narrative actions are equally loose and fluid. The goal, as established by Montaigne, is to embody a mind actively engaging its work:

I go out of my way, but rather by license than carelessness. My ideas follow one another, but sometimes it is from a distance, and look at each other, but with a sidelong glance. […] It is the inattentive reader who loses my subject, not I. Some word about it will always be found in a corner, which will not fail to be sufficient, though it takes little room. I seek out change indiscriminately and tumultuously. My style and my mind alike go roaming. (“Of Vanity,” 925)

Central to the essay’s action, then, is an associative framework filled with digressions, expansions, references, and generally an intentional (‘licensed rather than careless’) lack of linear progression. Addressing this formal style as an epistemological activity, Adorno notes that “[i]n the essay, concepts do not build a continuum of operations, thought does not advance in a single direction, rather the aspects of the argument interweave as in a carpet. The fruitfulness of the thoughts depends on the density of this texture. Actually, the thinker does not think, but rather transforms himself into an arena of intellectual experience, without simplifying it” (Adorno [1958] 1984: 160–161).

Adorno’s image of an arena adds weight to the sense of essay as narration, because it implies the ‘transformation’ of the essay as a static conveyance of thoughts into the active behavior of an essayist marshalling ideas and experiences that are unfolding before rather than being pursued (let alone caught) by the reader. The resulting sense of drama and temporality are in keeping with Montaigne’s sense of the form, and also with a sense of his persona. That overall sense enlivens the elements of Kellogg and Scholes’s classic definition of literary narrative: here there is the vital “presence of a story and a story-teller.” Such fusion is often lacking in nonfiction films whose voice-of-god narration is the essence of extradiegetic voiceover. But the essay’s persona is part and parcel of the story. That is, the story of the essay is the immediately present persona working through the ideas and issues at hand. It is the ‘drama’ or tension of the thinking, embodied in the figure of the persona, that makes for the drama and tension of the narrating. Like Scheherazade, the essayistic narrator does not resolve the essay situation, but merely ends the current discussion.

This action, central to the essay, can be more immediately recognized in the film essay because it is more deeply integrated and immediate in that form. As the film essay pursues its own desire to use oral narration and yet avoid voice-of-god omniscience, it is precisely a refusal of narrative fixity and authority, as well as an ostensible division between teller and tale, that the essay film put forward – to the extent that many essay films literally place the figure of the narrator/filmmaker in the frame – not to increase authority but to make the narrator an element of the cinematic. The ensuing fluidity of persona, coupled with a denial of certainty in the essayistic voice, creates the foundation upon which the film essay is constructed. ^[5]

This particular alignment of the cinematic with the essayistic is especially fruitful, since it provides an opportunity to double – at a minimum – Montaigne’s prose-based multivocality through visual narration. This possibility is a function of what Verstraten describes as the creation of a “fictive narrator” that is actually made up of “a visual and an auditive narrator” (2009: 10). By playing with these and other permutations available in the medium, the film essay becomes not simply a multiplication of representational realities, but an excess of narratives, voicings, and focalizations, all of which serve to jam the machinery of documentary’s traditional narrative certainty and proclamation. The result is a shifting of the nonfiction film essay’s relation to ‘the real’, converting it from predominantly mimetic representation to enactment of narrativity as performance in a complex and highly playful embodiment of the nature of narrative itself.

4. Narration and the film essay

With that framework in mind, most characterizations of the essay film trace its heritage back to that of Montaigne’s prose essay, his declared stance of “Que sais-je?,” and the form’s overall rejection of anything representing a voice-of-god declaration of authority. ^[6] Instead of omniscient, voice-of-god conviction, within the Montaignean prose essay we find the ostensible voice of a lone author, and voices of multiple other authors via paraphrase and citation, and myriad examples and allusions, and a plethora of digressions and side-steps away from the principal topic – all in service to a fluid persona. For its part, the film essay is often an intentional response to the auditory narrative voice in nonfiction film that refuses to go away. Essay films often use oral narration to directly evidence a particular desire to turn traditional (or at least popularly conceived) spoken narrative on its head, to use the tradition in ways that question the assumed authority of narrative voice via reflexivity. In short, the film essay provides a prime example of metafilmic activity in the cinematic process, more specifically in service to a de-authorizing of voice-of-god narration. Ross McElwee’s Sherman’s March demonstrates such questioning of traditional documentary’s narrative performance with a simple play on the voice-of-god certainty that we witnessed earlier in Capra’s propagandistic World War II documentary. It does so through its opening, where the film clearly echoes the didacticism of classic ‘World at War’ propaganda.

On its face, this opening is precisely what so many expect from – and so many dislike about – traditional narration in nonfiction film. It is a singular voice, distant in its voiceover, declarative in its objective rendering of reality. Hardly what one would expect form a film essay devoted to decentering narrative voice. But the scene is not yet over, as the voiceover on the still of Sherman becomes voice-off in the ensuing seconds:

This commentary on the voiceover, this voiced annunciation of the voiceover’s construction, re-voices the narrative. It is not often, after all, that we get to hear the voice of god editing himself. The mimicry of the scene signals McElwee’s awareness of the narrative didacticism so often and readily ascribed to documentary. Voice-of-god narration, and the equivalently didactic formula of graphic illustration, is followed by the narrator’s – now off-script as well as voice-off – comments and the subsequent dialogue between McElwee and his narrator/former teacher, Richard Leacock. What starts as of a dutiful following of the rules becomes a parodic breaking of omniscient, voice-of-god narration –and even of the ‘omniscient camera narration’ of direct cinema.^[7]

While McElwee provides us with a clear – and clearly oppositional –response to traditional voice-of-god narration, such opposition to documentary certainty is not new to the film essay. It exists as early as Bazin’s ostensible naming of the genre in his review of Chris Marker’s ground-breaking Letter from Siberia. McElwee echoes Marker in questioning the act of narration – but here not through a self-questioning, voice-of-god narration as in Marker, but through what now appears almost as voice-off narration, a sense of the narration as immediate performance. Such narration already moves beyond that established years earlier by Marker, and McElwee will quickly the complicate the process of narration further.

Here we have not simply a voiceover questioning its own narrative, but apparently an introduction of the narrator himself. Viewers have no direct proof of this, of course. But they are likely to infer, via the normal grammar of film, that the figure we see sweeping the floor in an abandoned apartment is the figure who narrates his presence for us. After all, it does not take much acquaintance with visual narration to unite a figure who narrates losing his girlfriend and uncertainty over his future, with a figure cinematically framed in deep space who is aimlessly sweeping the floors of a desolate, unfinished, and abandoned apartment. I = he, as it were, in a Proustian collapse of first- and third-person narration into a single narrative figure who narrates both orally and visually, even as the ‘information tracks’ remain somewhat separate.

This brief sequence begins to suggest the capacities of film to address and extend the activity of the Montaignean homodiegetic narrator. Filmmakers have not missed those possibilities, and the ensuing interest in Montaigne among film essay theorists is deep-seated and not limited to the idea of undermining voice-of-god narrative certainty. There is an equivalent interest in Montaigne’s declaration that he takes “himself” as the material for his essays, in his desire to use the medium of writing to understand himself in the act of understanding others and other issues. At the same time, while the author is implied in the matter of the essay, it is through a persona that the essay enacts thinking lest it become memoir. As the Corrigan epigraph emphasizes, it is the act of presenting the narrator thinking within the medium of representation that defines the essay .^[8]

While the appearance of that dynamic is often historically tied to Marker, there are multiple precursors to be found, in regard to both cinematic questioning of one’s own narration (e.g., Nuit et brouillard) and linkages between essayistic composition and cinematic production (e.g., Astruc’s concept of caméra-stylo). As for the latter, although damaged by its linkage to heavily romanticized concepts of auteurism, the camera as pen finds new echoes in the practice of today’s film essayists. For example, it is closely echoed in Agnès Varda’s concept of “cinécriture.” As Brioude notes, “Le concept de cinécriture, inventé par Agnès Varda elle-même, signifie ce qui est écrit avec la caméra, son ‘phrasé’ spécifique pour le distinguer de l’écriture littéraire qui requiert la plume” (Brioude 2007).  It is not so much a distinguishing from, however, as a comparison to. In Cruikshank’s words, “Varda describes her filmmaking – both documentary and feature – as cinécriture: “En écriture c’est le style. Au cinéma, c’est le cinécriture” (Varda in Cruikshank 2007: 120). ^[9]

However écriture and cinécriture are parsed, Varda’s documentary film practice is fairly easily aligned with the written essay, both in realization and conceptualization. She herself describes her Les Glaneurs as a “petit documentaire d’art et essai ” (Varda in Cruikshank 2007: 129), and it might well serve as a model of the form, even more than McElwee’s heavily autodiegetic March, whose full title is indicative of the dual plot lines running through it: 1) creating a documentary of Sherman’s March while 2) Meditat[ing on]the possibility of romantic love […] during an age of nuclear weapons proliferation. To a degree, the film drifts toward autobiography, although it mainly draws on events from life to further the narrative. For its part, Varda’s film is more closely centered on and through the tradition of a focused topic (“On Gleaning” would be the classic way of titling the piece), which is then followed out through the again traditional elements of example, digression, expert citation and so on. Enunciating that focus, Les Glaneurs devotes both visual and vocal narration to revealing the narrator in the process of working through the question of gleaning, as the following scene makes clear:

Together, the voiceover and the visuals narrate the performance of topic and relation of self to topic that is occurring on-screen. Varda metaphorically suggests (visual narration) and directly states (voiceover) that the subject of her essay, gleaning, is more than a means of moving into other, more personal topics. Echoing the Montaignean form, then, Les Glaneurs does display a marked proclivity for pursuing self through direct reference, and counter-reference, and self-reference. At the same time, it directly narrates connections between the activity of gleaning and the process of filming, intermixing oral and visual narration, providing the active and actively performed intertwining of self and subject central to the essay. As announced in the French title, the film pushes the boundaries of narrative and narrating as the subject of the camera and subject of the essay co-occur in many places.

At the same time, Les Glaneurs also maintains some distance between narrator and narrated. Although Varda is a constant presence in the film, both aurally and visually, the two forms are not equally distributed and their function may remain relatively separate. For one thing, Varda is heard primarily in voiceover and at times in actual voice-off, thus maintaining standard documentary narrative form as far as the voiced (i.e., not on-camera) narration goes. When she is fully in frame (rather than fragmented into a body part, such as her hand), Varda remains largely ‘voiceless’. To some extent, this separation is underscored by the way in which her presence is often noted as a form of metalepsis or intrusion into the diegesis. When fully visible, she is present more as a subject for perusal – “Agnès Varda” – than as the active narrator. As Pethő notes, these appearances may operate less as outside intrusions into a fictive diegesis and more as a negotiation between Varda’s ‘worlds’: “the reality of herself, the personal world of the author-narrator and the reality captured by cinema vérité style cinematography” (Pethő 2010: 69).

Ostensibly similar metaleptic intrusions may differ, then, in the way they affect the narrative and the way they are experienced. Often, Varda (or more accurately, part of Varda – hands, hair) may be seen at the same time that she is heard in a way that suggests the immediate time of the film’s narration. In such instances, when she speaks she is ‘of the moment’, and her voiced narration presence matches more readily with her physical visibility. As Varda discusses her hands, for example, there is little sense of a gap between the space and time of the visual narrative and the space and time of the spoken narration. Because of the autodiegetic components of the film, even when focused on ‘abstract’ topics such as gleaning, the voiceover narration paired with the visual presence is less readily experienced as metaleptic (i.e., the narration may be experienced as voice-off within the diegesis even when it is actually post-production voiceover). 

But when Varda is visually within the frame, a sense of separation can be felt by the viewer, as Varda is visually converted from speaker to subject, especially when there is what is clearly voiceover narration. The ensuing sense of observation or even mild voyeurism suggests a gap between Varda as narrator and Varda as subject that is likewise suggestive of a temporal gap – and of the temporal shiftings of written essay in turn, a gap that may more readily fall into the metaleptic for many viewers, although it may not be clearly intended as such by Varda. As was mentioned above, the prose essay works to suggest the immediacy of thinking, while at the same time giving free rein to memory and reenactment as it plays with the ‘location’ of its narrator in space and time. Varda’s presentation of herself echoes that play with time and narration, regularly easing the metalepsis to produce instead a satisfying complexity of self-analytic persona well within the bounds of the traditional essay form.

Of course the essay, as characterized by Adorno, is a radical form unwilling to stay within the bounds of its own tradition. Not surprisingly, then, it is not hard to find further plays with narrative and narrator. In McElwee, for example, the roles of implied author/filmmaker, narrator, and on-screen performer can appear to be completely contained within the immediate frame of the autodiegesis, as the following clip shows:

Within this scene, there appears to be almost no complication in narrative positioning. The immediacy of action suggests an almost complete erasure of ‘external’ focalization, since the camera is visually referenced as being under McElwee’s control, thus suggesting a reduction of that device to mere mimetic recorder. The immediacy of the narrator’s eye-to-eye exchange with the viewer tends to override viewers’ awareness of the presence of camera or any externality whatsoever. But while the scene is subject to the agency and control of the implied author, not to mention the filmmaker, the entire narrating process appears to be contained within the frame, to a greater degree than is found in Varda’s ostensible voice-off narration of her fragmented appearances or her narration of her observed self.

It would be surprising, of course, if McElwee did not choose to undercut this narrative situation as well, and he does so in a very direct way in a later film, Time Indefinite. Reprising the scene just discussed, McElwee intensifies his essayistic play with narration in a way that reflects not on self alone but also on the process of self-reflective narrating.

McElwee complicates this similar scene’s seeming mimesis and relatively straightforward focalization through his direct voiceover commentary on his on-screen narration. While that extradiegetic complication ostensibly critiques his intradiegetic narration, it achieves much more. His shifting of a fully-contained focalization through an extradiegetic narrative echoes a movement readers have been observing for a long time in fiction, of course, through direct address of the ‘dear reader’. What is intriguing here is that the break is not complete, a function of what Chatman describes as “cinema’s access to not one but two information tracks – sight and sound” (1999: 315). In this case, the sound track is itself doubled, containing at the same moment not only McElwee’s diegetic statements but those of his voiceover narrative as well. The viewer is placed in the midst of an immediate performance, doubling rather than metaleptically breaking, as it were, the process of essaying the self that forms the heart of the genre. The narrative action is both reflectivity and reflexivity in active union, internal and metatextual, with McElwee’s persona moving in and out of its various roles, none of which fully disappear or are completely broken – a feat extremely difficult, if not impossible to achieve in written narration.

Such narrative play can be and is complicated in other ways by McElwee. Not content with his pursuit of off-screen narration as a means of questioning the facticity of documentary practice, McElwee is willing to put the role of narrator itself up for control. In the opening of the following sequence from Bright Leaves, we are presented with an ostensibly untroubled, mimetic offering of a standard street scene via the camera’s slow, downward pan and tilt. Given the absence of any auditory narration, we are apparently free to read this as direct or pure representation. But as we will see, the unfolding of the scene begins to trouble the narration in ways that seem tailor-made for discussing the issue of focalization.

Even the most direct or pure representation has several qualifications that occur in the interaction between scene and viewer. There is the visual narration/focalization provided by the camera apparatus. But that ‘naïve’ narration is qualified by the now fully established awareness of McElwee as the implied/assumed camera operator – a presence that, given the direct somatic reaction of the viewer to camera movement, is likely to be more strongly sensed than the production activities associated with ‘written’ or print media. The viewer is likely then, given all that has occurred, to see this not simply as a capturing of the scene but a capturing of the scene by Ross McElwee. He is, by now, established as agent within (behind?) all aspects of this filmic production, the ultimate focalizer who drifts toward a diegetic embodiment of the implied author. Even with this apparently simple mimetic representation, then, we already we have complex focalization at work.

It might be argued, of course, that the absence of voiceover or voice-off narration, combined with a fully sutured viewer, erases this sense of complexity, leaving behind not a somatic connection to McElwee but the sense of simple, mimetic camera narration. But that possibility is broken by the intrusion of McElwee’s auditory narration, as he declares in voice-off: “I’m in Winston-Salem, North Carolina.” With that, the camera narration quickly expands into the intradiegetic, homodiegetic, focalizing narration of McElwee. 

Paradoxically, such a narrative stance also suggests simplification or at least identification, a return to a singular, declarative mood. In direct perceptual terms, we can now consciously merge the literal, somatic focalization of the camera with the narrating focalization of McElwee. Such positions of assurance are always suspect in the essay, of course, and certainly within that of McElwee’s film essays, with their continual vibration or tension between the resting moment of a fixed narrative stance and McElwee’s relentless undercutting of that position. Immediately, then, McElwee qualifies his own voice-off resolution of the scene that he and camera narrate. This is not a pure representation via camera, nor is the voice-off a declarative mood. Instead, the scene is re-narrated in a way that shifts the relation between narrator and narrating into yet another mood.

This would be Winston-Salem if McElwee’s initial statement were true or if the camera did not lie. But cameras do lie, or at least they are not the pure mimesis assumed by popular ideology and furthered by classic Hollywood style. “This is not really Winston-Salem.” It is not even a two-dimensional, camera-based representation of Winston-Salem. As McElwee’s now-perhaps voice-over further states, this is a simulation of the city, a “permanent film set” used by students at the North Carolina School for the Arts, and McElwee is here not to film the cityscape but to interview Vlada Petrić, “the noted film theorist and historian.”

Why this shifting of McElwee’s relationship to his own statements? Why this deliberate reworking and undercutting of his own visual and vocal narration? Simply answered, to undercut his own visual and vocal narration. That is, to rapidly pass back and forth over the gaps between implied author, unreliable narrator, and intradiegetic narrator so as to bring forward narrating itself, be it visual or oral, and to signal both its presence and its very fluidity. This vignette thus serves to repeat and instantiate the Heraclitean stance that has been a part of essay behavior since the time of Montaigne. In these few short narrating moves, McElwee shifts positions and effectively undercuts the comfortably declarative mood of voiceover and voice-off narrating in the process. 

If this scene offers a prime example of shifting the relation between narrator and narrative statement, its continuation offers a prime example of how to enact the fluidity of the essay through playing with multiple layers of focalization and/or embedding. Having announced his purpose in coming to this set – to meet with Vlada Petrić – McElwee cuts directly to Petrić himself, effectively handing over the vocal narration to Petrić while maintaining control over the visual narration via the camera. In essence, there are now two focalizations present, that of McElwee/camera and that of the figure in the frame, Petrić. Importantly, the cut to Petrić is immediately paired with him speaking directly to the camera, thus controlling the audio narration in a direct look at the viewer. Technically we are looking through McElwee’s eyes (as ‘they’ look through the camera lens). But Petrić is determined to control the scene himself.

Not only does Petrić appear visually at this point, but his dialogue also is the first clearly within the diegesis of the scene. Petrić speaks in dialogue and in the same temporal moment as his appearance. With this shift, we enter the world of cinema as we normally experience it. We are viewing the unfolding of narrative before us, with characters and story fully inhabiting this time and this action. Interestingly, however, the narrating is itself ambiguous in reference and in time. “You can use it or you can discard it,” Petrić declares. “If it’s not good you can discard it.” Here is a bit of temporal qualification. The viewer must guess as to what ‘it’ refers, inferring that the reference is to something stated at a previous time, the time between the shooting of the street scene (at least in the story – that scene could have been shot later) and the shooting of the Petrić entrance scene.

Of course, the reference is to this footage itself, to the event now unfolding. But the comment is also one about whether or not to use that footage. The discussion is one of how the narrative is to be constructed, or how eventually the story is to be narrated. This is important, because what it raises once again is the reflexive issue of this narrative as constructed, with all that implies. Finally, there is also visual ambiguity added to this mix. What is this wheelchair doing in the shot? Why is Petrić duct-taping wooden slats to the arms? Why are we being visually ‘told’ of Petrić’s interest in this prop? Eventually the visual narration answers that question. It is the classic tracking/dolly shot for films on a budget: McElwee is placed into the chair (we are told this via the position of the camera in relation to Petrić, by fragmented shots of McElwee’s pants legs, and by Petrić telling McElwee to get into the chair). Finally, Petrić begins to enact his lecture on film and kinesthesia by pushing McElwee and camera around the set.

What ensues is a hilarious tossing back and forth of narrative emphasis, with McElwee, camera, and Petrić vying for control of the situation and its storyline. The alternation in the dialogue between McElwee and Petrić tends to suggest a sharing of focalization (in addition to that provided through the camera). Yet the topic itself, the 1950 Michael Curtiz film Bright Leaf, is suggestive of a further duel over storyline, since the question addresses McElwee’s basic pursuit in this film essay: that the story Bright Leaf is the story of his family. To discuss that topic, then, is to discuss how the plot line is to proceed. This metatextual activity is part of the overall struggle for control, especially since McElwee has, to all surface intents and purposes, had the oral narration effectively yanked from him. In addition, although McElwee never relinquishes camera narration, that control actually serves to instantiate his subordinate position. He is seated on a lower plane than Petrić, he is being wheeled backwards around the film set, and he does not even occupy a visible presence in the shot – all that is given over to the looming Petrić, who continually dismisses McElwee’s feeble attempts to reintroduce his status as narrator and prime focalizer in the narrative.

Having used the scene to play with focalization and to engage the question of narrative control, McElwee finally reinstates both. Interestingly, he does not do so visually – rising from the chair, cutting to another scene, or so on – but aurally, shifting the narration out of the frame per se and into his own thoughts as heard in near voice off. This choice is interesting on two key grounds. First, it signals viewers’ departure from their familiar, comfortable role of inhabiting a present space as it unfolds before them. Where does this narrating come from, after all? Second, this shift echoes that earlier voice-off narration prior to Petrić’s entrance and the audience’s immersion into the cinematic now (“I am in Winston-Salem”), a scene already used to signal narrative uncertainty. When McElwee wrests control of the narrative focalization from Petrić, the tension is seemingly resolved. But the film’s and its persona’s lack of true voice-or-god dominance is only reinforced by the need for such extradiegetic narration.

This tug-of-war with focalization as a means of playing with essayistic narration is furthered in McElwee’s related play with narrative time, a playing that extends beyond the ambiguity of reference noted above. “I am in Winston-Salem,” McElwee declares. But of course he is not, and not only because he is on a film set. The narrating McElwee is in the sound booth, providing a subsequent narration added after the initial filming. This time gap is not as readily apparent as with the shift in focalization via the cut to Vlada Petrić. But it is underscored by the audience’s movement out of the immediacy of the cinematic moment. While we may ostensibly be occupying the narrative thoughts of McElwee, there remains an awareness that these thoughts originate not in the diegetic moment but in a later moment now brought into this one, a temporal shift that is part of Mary Ann Doane’s (1980) sense of the “uncanny” nature of voice-off and voiceover, an uncanniness not only of disembodiment but of temporality.

The overall play within this scene thus serves to reinforce the nature of the essayistic narrator itself as the activity of “essaying” in the Montaignean sense: a fluid attempt to narrate a sense of self in relation to the materials that make up the essaying. McElwee signals that goal in the opening scene of Sherman’s March, and he thoroughly enacts it this scene from Bright Leaves, weaving in and out of the topic through complex play with the narrating voice – a regular, even dominant, component of his films. McElwee can do so while maintaining an effective story line because the essay film lends itself to such embedding, echoing, and expanding upon what has been a primary process of the prose essay since Montaigne. Such expansion is not problematized but enhanced by the medium of film, with its multiple venues for providing narration, both visual and auditory. 

There is one further component of the prose essay that has been alluded to at various points above: the metatextual. McElwee’s parody on traditional oral narrative and Varda’s interest in aligning her filming with gleaning hints at such metatextual commentary, without dramatically breaking storyline. McElwee parodies voiceover narration to contrast it to the voiced narration of his film. Varda mimics the act of gleaning by filming her hand closing around a truck, using voiced and visual narration to connect theme or topic in the diegetic and her own activity as filmmaker. Such reflecting on the narrator’s behavior in relation to the ostensible topic at hand can be found throughout the prose essay. But there is a further form of reflecting that entails direct discussion of the medium itself, a form not of reflective so much as reflexive analysis that has a long history in both genres. 

An often cited example from Montaigne would be the following, from his essay “On Repentance”: “my book and I go hand in hand together. Elsewhere men may commend or censure the work, without reference to the workman; here they cannot: who touches the one, touches the other.” Montaigne subsequently will link the theme of repenting to the issue of earlier self-representations. But this and lengthier elaborations extend beyond thematics to enter into a concern with the medium itself. Montaigne has made such commentary central to the essay form, and McElwee is also quite ready to introduce similar moments into his films. Here is an example from Bright Leaves:

Here there is a stronger metaleptic intrusion into the diegesis than that offered at other times. McElwee narrates his thoughts on his relation to the cinematic medium itself, and in doing so appears to more fully leave the diegesis. But the essay form encourages and subsumes such digressions, enveloping them within the overall act of essaying, rather than establishing them as full ruptures with the narrative progression. It is a dynamic enacted in Varda’s Les Glaneurs, as well, as the following early scene makes clear:

In discussing the making of Les Glaneurs, Varda has commented on her filmmaking in a way that establishes why these digressions, while more metaleptic than those reinforcing narrative themes, are still subsumed within the boundaries of a particular essay. In an interview on the film, she refers to the dual possibility of both appearing in frame as performer while remaining outside as camera/author. “I have two hands. One has a camera – the other one is acting, in a way. […] I like very much the idea of the hands. The hands are the tool of the gleaners, you know. Hands are the tool of the painter, the artist” (Varda 2001).

It is a striking image, because it suggests the unification of several levels of narration within the figure of the narrator. Hands glean, paint, film and, through film artistry – with its capacity for immediacy of appearance and voiceover – the film essayist appears to have found a way to solve Virginia Woolf’s paradox lying at the heart of the prose essay: “Never to be yourself and yet always.” In Varda’s cinematic image, self and narration seem to unite. Of course that is a trick of film’s sense of the immediate, as provided by the multiple tracks allowed by visual and auditory narration. Yet for Varda, especially given her orientation toward performance arts, Woolf’s paradox in writing is not a problem in film; it is an opportunity, an opportunity to enact multiple selves through multiple narrative positions or levels within a highly malleable medium. In a postmodern world, never to be oneself is to be oneself, and the act of narration is always an ongoing dynamic, a desirable fluidity rather than a vacillation between poles.^[10]

The possibility of myriad narrative modes and embeddings is a benefit of cinematic narration’s complexity, and both McElwee and Varda celebrate it, often visually enacting their belief with plays on multiple visual selves via mirrors. We have seen this sort of cinematic embedding in McElwee’s motel scene, and Varda uses it extensively in Les Glaneurs as well. But perhaps the most specific acknowledgment of this embedding can be found in the close to Varda’s Beaches.

The sequence has a variety of metaphors at work, both directly within the diegesis and as part of the visual narration. The 80 brooms are a metaphor for Varda’s 80 years. At the same time, the dolly shot out, with its revealing of additional brooms as it moves away from Varda, adds a spatio/temporal progression to the metaphor, that of passage through the years. Finally, there is the mirroring of the scene within itself, mimetically in the photo of brooms that Varda holds and then metaphorically in the literal mirroring of Varda herself now holding a mirror in place of the original still of brooms, and so on. In this multi-level narration, all visual, Varda provides one last image of the essay, in which the diegesis is not simply reflecting but also encompassing the extradiegetic world through a play with focalization and narrative levels theoretically capable of infinite expansion.

This is the ultimate essayistic position, of course, with its suggestion of infinite regression or expansion – infinite embedding – even without the presence of spoken narration. Adding that information track would only add, if possible, to the infinite play that is possible. As such, the scene makes two key points to nicely encapsulate this overall discussion. First, the play of visual narration in the film essay is immensely complex, operating in ways that continually problematize the idea of – and desire for – documentary as simple mimetic representation. Second, voiced narration, once an annoying simplification and destruction of the visual power of cinema, is now happily its opposite – a complication of narrative levels and possibilities worthy of equally complex study and critical play. The next set of variations, and they will appear, should only add to the conversation.

My thanks to Rebecca Matthews and Dustin Zemel for their help in completing this article.

---

References

---

Notes