Rafe McGregor

The purpose of this article is to examine the ontological effects of digital 

technology, and determine whether digital films, traditional films, and pretraditional 

motion pictures belong to the same category. I begin by defining 

the parameters of my inquiry, and then consider the two most significant 

consequences of the new technology. §2 proposes a decisive refutation of 

the causal relationship between reality and photography. §3 identifies an 

end to the dominance of photorealistic film over animation, and argues for 

an inversion of that relationship, whereby animation is paramount. Finally, 

I consider the implications of these consequences for film ontology, 

compare theories, and conclude in favour of Berys Gaut, for whom digital 

film is the latest incarnation of a history of moving pictures that stretches 

back for centuries. 

Defining the Question: Terminology and Technology 

The fundamental problem with the question ‘what is film’ is that it 

presupposes a certain kind of answer. In fact, the words ‘film’, ‘motion 

picture’, and ‘movie’ all make the question redundant. The term ‘film’ is 

derived from film stock, the photosensitive chemical material used to record 

rebounding light. The conversion of light into film produces a negative, 

which is then photographed itself to produce an original print. The 

photochemical process has been labelled ‘traditional’, and produces both 

still photographs and the sequences of still images called ‘films’. So it 

seems that asking what film is limits the answer to the projection of 

photographic images onto screens, exhibited through the medium of cinema 

from 1895. Although photographic film is still used, the success of George 

Lucas’s Star Wars: Episode II – Attack of the Clones in 2002 heralded the 

entry of digital cinema to the mainstream market. James Cameron’s Avatar 

(2009), a three-dimensional (3D) digital production, is the most 

commercially successful film to date at the time of writing, and a return to 

photographic film in the future appears highly unlikely. 

The use of ‘motion picture’ and ‘movie’ is no less prescriptive. The 

etymology of ‘movie’ is uncertain, but is probably an abbreviation for 

‘moving image’ or ‘moving pictures’. I shall regard ‘moving image’, 

‘moving pictures’, and ‘motion picture’ as synonymous for my purposes. If 

the inquiry concerns moving pictures rather than film, then it extends back 

well beyond 1895. Gaut (2010: 6-10) identifies three different categories of 

motion picture: object-generated, handmade, and mechanically-generated. 

Object-generated motion pictures have existed since at least the tenth 

century CE, in the form of Indonesian and Chinese shadow puppet plays. 

Handmade motion pictures include simple flip-books, magic lanterns, the 

thaumatrope, the phenatakistocope (1832), the zoetrope (1833), the 

choreutoscope (1866), and the praxinoscope (1877). Mechanically-

generated motion pictures began with Edison’s kinetoscope and the Lumière 

cinématographe – both of which used film stock – and the category includes 

subsequent electronic and digital developments. 

While Gaut maintains that all three types of moving image share a 

common ontology, Noël Carroll (1996: 127-130) restricts motion pictures to 

the mechanically-generated. I wish to avoid asking a redundant question 

and limiting the scope of film or motion pictures without justification. I 

shall therefore use ‘film’ and ‘motion picture’ interchangeably, with both 

referring to the exhibition of moving images in the broadest possible sense, 

i.e. not restricted to either photochemical stock or photorealistic pictures. I 

include animated features in the category of traditional film due to the fact 

that until Toy Story (John Lasseter, 1995) animated films consisted of 

photographs of either illustrations or objects (for stop motion animation, e.g. 

Wallace and Gromit [Nick Park & Nick Rushton, 1989-2010]). 

Carroll (1996: 130) distinguishes film from similar art forms like 

theatre and puppet shows by stipulating the moving image as two-

dimensional (2D). With the post-Avatar revival of 3D film, this distinction 

might seem outdated, but 3D cinema has existed since the screening of The 

Power of Love in 1922, so the issue is neither new nor linked to digital 

cinema (Zone 2007: 1-2). I endorse Carroll’s distinction, specifying that 

motion pictures are 2D even though they may appear 3D. The images are 

generated on a two-dimensional surface; they may be drawn on this surface 

directly, in the case of a flip-book or a zoetrope, or projected onto a flat 

screen, in the case of a shadow play or traditional film. I can now identify 

the question, which is ‘what is film’ where ‘film’ is the art form involving 

the exhibition of 2D moving pictures, and ‘a film’ or ‘motion picture’ is a 

specific instance of this category. I have noted the possible beginnings of 

the art as shadow plays, mentioned the more recent precursors to traditional 

film, and given a very brief description of the photochemical process. A 

summary of the digital revolution in film follows. 

In contrast with the photographic process, where light is converted 

to film, digital image recording converts light into streams of binary 

numbers. Digital recording employs a charge-coupled device (CCD) to first 

convert light levels to voltages and then convert them to the number streams, 

which are stored in bitmaps, i.e. grids of pixels. 2 The bitmap is a 

mathematical representation and thus has no physical relation to the image, 

which is manufactured by the interpretation of numbers. The images can be 

recorded either on digital tape or as digital files. Aside from the use of a 

CCD, there are two other methods of producing digital images: by hand 

with a software tool, and computer synthesis. The former creates images by 

digital painting; the latter involves the construction of a vector graphic or 

3D model to generate bitmaps. All three methods of manufacture can be 

combined seamlessly. The digital image does not suffer degradation when 

it is replicated, and duplication produces a clone rather than a copy (Gaut 

2010: 12-16). 

‘Computer-generated imagery’ (CGI) describes the application of 

pre-and post-production digital painting and computer synthesis to film, 

and is fundamentally ‘a form of computer animation’ (Cook 2004: 882). 

Steven Spielberg’s Jurassic Park (1993) was the first extensive use of CGI 

in a photorealistic film, and met with unprecedented commercial success 

(Cook 2004: 890-891). Two years later Toy Story became the first digital 

feature film, i.e. the first film created entirely with CGI. Star Wars: 

Episode I – The Phantom Menace (George Lucas, 1999) made even more 

use of CGI than Jurassic Park, and was exhibited in digital screenings in 

addition to its traditional cinema release (Gaut 2010: 10-11). The success of 

the sequel in 2002 ushered in a new era of motion pictures. The 

philosophical implications of digital cinema are both far-reaching and 

complex, but I believe it has finally quashed the various positions that reject 

photography as a representational art. 

Another Long Goodbye: Photography as Presentation 

Gregory Currie (1993) bade a conclusive farewell to the persistent – but 

insubstantial – hypothesis that the ontology of film was that it constituted a 

particular kind of language. The argument he calls the ‘presentation thesis’ 

has been just as persistent, with as little substance (1995: 48). Currie 

describes the presentation thesis as the claim that photography presents 

rather than represents the world, and focuses on Kendall Walton’s 

transparency argument. Gaut (2010: 21-22) maintains that there are two 

distinct objections to photography as art within transparency – causal and 

reproductive – and although I disagree, I shall mention all four of the 

arguments opposing representation in photography. 

Andre Bazin held that photography was essentially objective, by 

which he meant that the photographic image was (the same as) the thing 

itself: ‘Photography enjoys a certain advantage in view of this transference 

of reality from the thing to its reproduction’ (Bazin and Gray 1960: 8). For 

Bazin, looking at an old family photograph was to see the deceased relatives 

themselves, rather than to see a representation thereof. Rudolf Arnheim 

interpreted Charles Baudelaire’s comments on photography imitating reality 

as the idea that photography ‘was nothing but a mechanical copy of nature’ 

(1974: 155). If this is indeed the case, then photography has no capacity to 

express artistic intention. Roger Scruton (1981: 578-598) contrasted 

photography with painting: where the latter stood in an intentional relation 

to its subject (due to the representational act), the former stood in a causal 

relation, and could not be considered a representational art. Looking at an 

object through a camera lens (or at the resultant photograph) was thus the 

same as looking at a mirror or looking through a telescope; it merely 

presented the reality. Finally, Walton argued that to see a photograph of X 

was to see X, while to see a painting of X was not: ‘Photographs are 

transparent. We see through them’ (1984: 251). This transparency causes 

photographs to put one in perceptual contact with the object photographed. 

I believe these four theses – objectivity, reproduction, causal 

relativity, and transparency – all present a single challenge to photography 

as representational art, and I use ‘presentation thesis’ to describe this 

challenge. I shall first show why the presentation thesis does not hold – 

even for traditional photography – and then show the effect of digital 

photography on the debate before considering the implications for film. It is 

worth noting an observation by Currie at the outset: most advocates of the 

presentation thesis have limited their claims to photographic images that 

have not been significantly altered subsequent to exposure (1995: 48). So 

the presentation thesis is subject to a qualification, and even its proponents 

do not accept that seeing (a photograph) is always believing (the reality). 

Scruton is even more restrictive, limiting his observations to a ‘logical ideal’ 

(1981: 578). One could argue that if photographs are not always objective, 

reproductions, causally related, or transparent, then there is a capacity (of 

some sort) for representation inherent in the medium, but the move is not 

necessary. 

Gaut (2010: 30-31) provides a conclusive and entertaining refutation 

of the presentation thesis, using the example of Jean Auguste Dominique 

Ingres’s painting, Madame Moitessier (1856). He identifies a number of 

ideas expressed by Ingres in the work, and then imagines that the woman 

depicted had been painted by Oscar Kokoshka and Lucian Freud. Gaut 

notes the differences in intention and appearance likely to have resulted, e.g., 

from skin as pure as marble (Ingres) to skin that looks like raw meat (Freud). 

He then extends the thought experiment to photography, suggesting that 

parallel differences in appearance would result if the subject was 

photographed by Sir Cecil Beaton, Julia Margaret Cameron, and Diane 

Arbus. The photograph perceived by the viewer would differ significantly 

in all three cases – just as the painting by each artist would differ. These 

differences are caused by the intentions of the photographers and the 

capacity of photography to express these intentions indicates that it is a 

representational medium. I believe Gaut’s thought experiment is sufficient 

to turn the tables on the presentation thesis. There is no justification for 

regarding all unaltered photographs as presentations of reality; many 

photographs are representations, so the presentation thesis should be 

confined to those depictions where the intention of the photographer is to 

present (rather than represent) an object. One might acknowledge that, e.g., 

photojournalists and crime scene photographers typically present reality, but 

even this claim is debatable. 

The appearance of Kevin Carter’s Pulitzer Prize-winning Starving 

Child in the Sudan in the New York Times in March 1993 resulted in public 

concern for the infant portrayed (Keller 1994). When readers saw the 

photograph they believed that there was a real starving girl, and that the 

event captured in the image – a vulture awaiting her demise – really 

happened at a certain time and place. The assumption was correct, and 

would not have been made if the picture printed in the newspaper had been 

an illustration, no matter how life-like. When one sees a photograph in a 

context like a newspaper one tends to think of it as presenting reality. There 

seem to be degrees of realism: from actual perception, to looking at a 

photograph, to looking at a painting. Thus looking at Carter’s photograph is 

not as horrific as actually being in the Sudan with the child and the vulture, 

but is more horrific than seeing a sketch of her. Walton (1984: 247-248) 

makes this point by contrasting the immediacy of Francisco Goya’s ‘Even 

Worse’ (1810), part of The Disasters of War, with Timothy H. O’Sullivan’s 

photograph, Death on a Misty Morning (1883). 

Walton has selected a particularly visceral portrayal of the aftermath 

of a battle, but his choice is unfortunate for his argument: O’Sullivan was 

believed to have moved the bodies of the corpses he photographed in order 

to create a more dramatic effect, i.e. to express his intentions in his 

representations of dead soldiers.3 While (most) photographs undoubtedly 

appear more real than (most) other pictures, even photojournalistic images 

do not necessarily present reality. Two of the most famous photographic 

images ever taken serve as examples. There is an ongoing debate about 

Robert Capa’s Falling Soldier (1936). The question is not only whether the 

image was posed, as I have suggested in Walton’s example, but where it 

was actually taken: on the battlefield as claimed, or at a safer position 

behind the front lines (Mitchell 1992: 40-43). When Dorothea Lange took 

the photograph Migrant Mother (1936) she failed to notice that her subject 

had grasped a tent pole in order to steady her sleeping child. The thumb and 

finger that appeared in the foreground of the image were subsequently 

edited out (Curtis 1986: 17-20). 

My point is that in two paradigmatic cases of photojournalism there 

is room for doubt as to the reality, and room for expression on the part of 

the photographer. William J. Mitchell mentions the response to an 

international incident between the United States and Libya in January 1989 

as an exemplary instance of photographic evidence being contested – 

contested on the very basis that it does not present reality (1992: 22-24).  

Hoaxes like the Cottingley Fairies (1917) and the surgeon’s photograph of 

the Loch Ness Monster (1934) reveal that the capacity for deception 

afforded by photographic evidence is not a recent development. I offer 

these examples as a supplement to Gaut’s argument, to show that there is 

always – and has always been – a possibility for representation in 

photography, even when it appears at its most transparent.4 If this is true of 

traditional photographs, then the opportunity for representation in digital 

photographs is even greater, due to the increased possibilities in editing. 

Gaut maintains that the digital image is correctly identified as: 

a mélange (or blended) image – that is, it can be produced by any of 

three distinct techniques and each technique may vary in the 

proportion it has in the making of a particular image (2010: 45). 

He claims that even if one restricts one’s inquiry to the digital photograph, 

the software packages accompanying digital cameras allow photographers 

of all proficiencies a wide range of editing options. Mitchell believes that 

digital images should never be called photographs, even if created with a 

digital camera, as they differ as much from a traditional photograph as a 

photograph from a painting (1992: 3). Gaut rejects this radical distinction 

on the basis of the similar generative methods employed in traditional and 

digital photography (2010: 49). I agree with his assessment, though I shall 

show that he underestimates the impact of his own falsification of the 

presentation thesis. 

I have considered the presentation thesis in some detail due its 

implications for the ontology of film. If photography could only present 

rather than represent reality, then photorealistic film would be similarly 

limited as an art form. Scruton (1981: 598-603) is convinced that a film 

presents a drama, and that while drama is a representative art, the filming or 

photographing thereof is not. He holds that photography hinders rather than 

assists dramatic representation, because of the limits it places on 

interpretation and its link to fantasy rather than the imagination. For 

Scruton, therefore, the ontology of film is that it is no more than 

photographed drama. Arnheim contemplated the consequences of the thesis 

for contrasting reasons, and was concerned about the inevitable and 

undesirable conclusion that there could be no artistic expression in film 

(1974: 155). With the presentation thesis refuted, however, film can take its 

rightful place as an art form. 5 The digital revolution has not only 

strengthened the argument for film as art, but also – I shall argue – indicated 

the category of art to which digital and traditional film belong. 

3. The Digital Film: Photorealistic Animation 

Gaut (2010: 14) defines digital cinema as ‘the medium of moving images 

generated by bitmaps’, and Lev Manovich (1995) identifies it as ‘a 

particular case of animation which uses live action footage as one of its 

many elements.’ They both note a number of essential differences between 

digital and traditional film (Manovich 1995; Gaut 2010: 17-18). All three 

types of digital image are subject to direct manipulation, with the result that 

the visual realism of photography has a plasticity that was formerly 

confined to animated film. In digital cinema, the traditional distinction 

between the ease of editing (a sequence) and the difficulty of image 

manipulation (special effects) has been eroded, and the spatial and temporal 

qualities of images can now be rearranged with equal ease. As there are no 

longer any aesthetic features linking film to film stock, digital cinema also 

has a purity from its physical origin. Once live action material is digitised, 

the existing relation to reality is lost as the digital images become raw 

material (grids of pixels) for manipulation. Digital technology has eroded 

the distinctions between creation and modification, production and post-

production, as every image – whatever its source – is processed through 

various computer programs before the final cut. Lastly, the speed of 

computer processing has opened up new possibilities for interactivity in 

digital cinema. 

Even the most cursory examination of the philosophy, theory, and 

criticism of film reveals an overwhelming bias in favour of photorealistic – 

as opposed to animated – films. Gaut states that the ‘philosophy of film has 

concentrated almost exclusively on traditional photographic images’ (2010: 

19). Manovich takes an even stronger line, describing the marginalisation 

as resulting in animation being ‘cinema’s bastard relative, its supplement, its 

shadow’ (1995). Works of reference on film frequently pay scant attention 

to animated features, and many philosophers exclude them from their 

definitions and ontologies of film. When the American Film Institute 

published a list of the hundred best American motion pictures made from 

1898 to 1998, it included only two animated films: Snow White and the 

Seven Dwarves (David Hand, et al, 1937), forty-ninth, and Fantasia (James 

Algar, et al., 1940), fifty-eighth. When the list was revised in 2007, it 

included Toy Story at a rather disappointing – given its significance in the 

history of film – ninety-ninth place, and at the expense of Fantasia (which 

was dropped) (American Film Institute). Animated films have nonetheless 

played a significant role in the film industry, from the commercial success 

of Snow White and the Seven Dwarves to the artistry of Beauty and the 

Beast (Gary Trousdale and Kirk Wise, 1991), and the groundbreaking 

technological advances of Toy Story and The Polar Express (Robert 

Zemeckis, 2004). 

Gaut (2010: 17) believes that digital technology has undermined the 

dominance of the photograph in film. The (literal) photographic essence of 

traditional film, the link between film and film stock, has been severed. 

Live action photography is now merely one of several methods – methods 

which are themselves frequently combined – of producing the moving 

pictures that comprise a particular film. Manovich (1995) notes that digital 

technology has introduced the possibility of a feature film of one hundred 

and twenty-nine thousand, six hundred frames that is both indistinguishable 

from a photorealistic film, and created entirely by digital painting. He 

maintains that the combination of live action material, painting, image 

processing, compositing, 2D and 3D animation that constitutes digital film 

has returned the moving image to its roots, which lie in the animation used 

in devices such as magic lanterns, zoetropes, and praxinoscopes. For 

Manovich, therefore, digital film has reversed the traditional relationship 

between photography and animation, with the latter now dominating. 

Perhaps digital film is best described by Gaut, as a mélange. 

Certainly my first thoughts on seeing a trailer for Avatar were, ‘what is it, a 

film or a cartoon?’6 There seem to be two equally valid answers: ‘both’, in 

that it was created with live actors (using motion capture) and digital 

painting; and ‘neither’, because it doesn’t fit into either traditional category. 

David A. Cook (2004: 921-926) presents an interesting study by comparing 

the representation of the Japanese attack on Pearl Harbour in From Here to 

Eternity (Fred Zinnemann, 1953) and Pearl Harbor (Michael Bay, 2001). 

The former relied heavily on the use of actual documentary footage from 

1941, but appeared much less realistic than the latter, which involved 

extensive use of CGI. Simulated reality, created by the combination of 

digital photography, digital painting, and computer synthesis, is thus more 

‘real’ than the photochemically-recorded reality. Cook believes that the two 

films cannot be compared by the same standards, and agrees with Manovich 

that the digital revolution has created a new aesthetic where film has 

become a type of animation, and where production is the first stage of post-

production (Cook 2004: 925-926). 

Cook uses ‘photorealistic animation’ to refer to animation that 

resembles photography in its attention to detail, but I believe the term 

appropriates the sense in which animation can replicate reality more 

successfully than photography (2004: 883). Film can no longer be 

differentiated from animation because the development of digital cinema 

has completely blurred the distinction. Peter Jackson deliberately used CGI 

techniques with a strong photographic basis for The Lord of the Rings 

trilogy (2001-2003). He successfully created an appearance of reality, yet 

one of the main characters – Gollum – is entirely computer-generated. 

Robert Zemeckis’s The Polar Express is an apparently clear case of an 

animated film, yet his paradigmatic use of motion capture brings an almost 

disturbing element of realism to the characters. Films like Avatar and Alice 

in Wonderland (Tim Burton, 2010) – their 3D aside – fall somewhere in 

between, and these examples of the cross-pollination of photography and 

painting show that the line between photorealism and animation is not only 

unclear, but irrelevant. The distinction may continue to be made for 

marketing purposes, but the two types of film have converged in a single 

category, digital film. Photorealism and animation have been replaced by 

photorealistic animation. 

Whether one regards the digital film as animated (Manovich) or 

mélange (Gaut), the photograph is obviously no longer as significant to the 

phenomenon as it once was. I believe that film has been freed from the 

restrictions imposed by photography, with the result that the scope for 

representation has increased exponentially. Drawing on Walton’s work 

(1970: 143-145), Gaut (2010: 37-38) elucidates the link between artistic 

expression and the perception of film and reality. Perceiving a film and 

perceiving reality diverge in a number of ways. Where the divergence is 

fixed, e.g. the lighting in the theatre and the 2D structure of the screen, one 

cannot regard it as communicating any artistic intention. Where divergence 

is variable, e.g. the shape of the image on the screen and the movement of 

the camera, one can attribute intention and expression. The point can be 

generalised to all art: when there is choice, there is a potential for the 

expression of meaning; and the greater the capacity for choice, the greater 

the capacity for expression. 

I have mentioned the increase in choice in the change from 

traditional to digital photography, and the case is even more extreme for the 

moving image. The bewildering number of options available to 

contemporary film directors means that they have increased opportunities 

for expressing their intentions, and it is this intentional relation to the 

subject that makes film a representational art. Gaut (2010: 49) holds that 

the digital revolution has resulted in a more compelling argument for film as 

art; indeed, Scruton’s claim that film is merely a recording of representative 

drama is simply implausible for the digital variety. Gaut also notes the 

intentionality of digital film in that it ‘possesses the possibility, in its non-

photographic modes, of creating expressive content that does not require the 

recording of any reality at all’ (2010: 50). I believe, however, that he 

understates the importance of digital developments. 

Although photographs continue to be used in film, they have become 

one of several kinds of mathematical representations that constitute the raw 

material from which the finished product is created. The photographic 

element of film is no longer any more significant than, e.g., computer 

synthesis, and Manovich has noted that even photorealism need not involve 

any photography. The fact that digital film does not require the recording of 

reality not only means that the art form has the potential to stand in a purely 

intentional relation to its subject, but that it actually does so. The reason can 

be found in Walton’s standard and variable properties, alluded to by Gaut. 

In photorealistic traditional film the use of photography was a standard 

(fixed) property, like the 2D screen, i.e. there was no other option available. 

In photorealistic digital film, the use of photographs has become a variable 

property. Live footage has become one of several options for the director, 

and the decision to use it has thus become an expression of intention. 

Digital film, like painting, is always intentional. A contemporary director 

who used only photographs (of either type) in a feature film would be 

exercising a meaningful choice, and that choice would convey his intention. 

The consequence of this intentionality is that digital film is indisputably a 

representational art. 

What is Film in the 21st Century? 

I began by defining the scope of my inquiry as the art form of moving 

pictures, where the movement may be actual or apparent,7 and where the 

picture – despite appearances – is understood in the ordinary sense of being 

presented on a flat surface. I have established that all four categories of 

traditional and digital photography and film offer possibilities for 

representation, and noted that the digital film is a development of traditional 

film characterised by what I have called ‘photorealistic animation’. I shall 

now answer the question ‘what is film’, taking into account both traditional 

and digital varieties, and consider the restrictions an ontology of film should 

impose in order to make an enlightening classification of the art form. 

Currie produced the first comprehensive philosophy of film in the 

analytic tradition with Image and Mind in 1995. Although the ontology of 

film he proposed has much to recommend it, Currie’s position on non-

photographic film is opaque. He cites animation as an example that ‘more 

distant causes’ of the cinematic experience need not be photographic, but 

subsequently excludes all reference to animated film (1995: 4). It seems 

that he is either concerned purely with photorealistic film or – at most – film 

produced with film stock. By linking film to photography, Currie is 

committed to the view that Toy Story – produced with CGI – is not a film, 

and that Toy Story and Snow White are therefore different phenomena. The 

position is clearly erroneous. 

Digital technology – in the form of CGI – has been applied to 

photochemical film over a period of at least forty years, with notable early 

examples like 2001: A Space Odyssey (Stanley Kubrick, 1968) and 

Westworld (Michael Crichton, 1973). CGI has been used incrementally, as 

technology developed, and even Toy Story was recorded onto celluloid for 

screening, and thus used photochemical material. There is some dispute 

over the first film produced and released without employing any 

photochemical process, but the success of Star Wars: Episode II – Attack of 

the Clones in 2002 popularised the fully-digital film (Gaut 2010: 11-12). It 

would be unfair to attribute the view that digital films are not films to Currie, 

and I think it much more likely that he simply failed to foresee the 

consequences of the digital revolution.8 

Carroll was quicker to realise the implications of the new technology: 

in 1996, he noted the advances made in electronic cinema and predicted the 

possibility of films being produced on CD-ROMs (1996: 115), and the 

future film as ‘digitally synthesized images’ (1996: 122). He believes that 

the domain of film is moving images, which seems a promising start in the 

light of subsequent developments. He stipulates ‘moving images’ as 

opposed to ‘moving pictures’ in order to include abstract, nonrepresentational, 

and non-objective film (1996: 126). Carroll’s ontology of 

moving images identifies five necessary (but not jointly sufficient) 

conditions: they must have the potential to appear to move, be a detached 

display, be 2D, and be generated by a template that is a token such that 

performance tokens are not works of art (1996: 130). That an image must 

have the potential to appear to move seems obvious. By ‘detached display’ 

Carroll means that the images must not contain egocentric information, i.e. 

tell the viewer where he is in relation to the space depicted.9 Thus watching 

Jurassic Park does not tell me how to find the park in question, nor identify 

my position in relation to the dinosaurs roaming on screen. I have already 

accepted the 2D requirement, which is implied in ‘picture’ (although not in 

‘image’). So far, therefore, Carroll’s ontology is applicable to digital and 

traditional film, and the range of motion pictures that preceded traditional 

film. 

Carroll (1996: 127-129) uses the type-token distinction to contrast 

film with theatre. The play Richard III is a type, and its performance is 

generated by an interpretation. I may, e.g., have the option of seeing two 

Richard IIIs, one at the university and one at a theatre in town. The 

interpretation of Shakespeare’s text will differ according to the tastes and 

skills of the different directors and actors involved. These interpretations 

are themselves types, and they generate the token performances, e.g. a week 

of six shows from Monday to Saturday at the university. These tokens will 

ideally be identical, but the two interpretations may differ substantially. In a 

play the text, the interpretation, and the performance are all subject to 

artistic evaluation. Richard Loncraine’s 1995 film Richard III differs 

entirely, however. I could have seen exactly the same performance no 

matter what film theatre I went to because the token – the particular 

screening I did in fact see – was generated by a template, which would have 

been a film print in 1995, and could be a film print or a digital file fifteen 

years later. The template is itself a token of the type, with the result that the 

performance I see is a token generated by another token. Neither the 

template nor the performance are subject to artistic evaluation, only the type, 

Loncraine’s Richard III. The interpretation, acting, cinematography, and 

other aesthetic elements of the film are all contained in the type. The 

template and the screening may be faulty, but such problems are technical 

rather than artistic. 

It should be evident that Carroll’s use of type and token 

differentiates plays and, e.g., puppet shows, from traditional and digital film, 

and is thus satisfactory in dealing with bitmap-generated motion pictures. 

Strangely, Carroll describes his ontology as ‘overly inclusive’ for ‘what we 

typically call motion pictures’ (1996: 131). He specifically excludes mass-

produced flip books and the zoetrope from the art form he is attempting to 

identify because they ‘do not seem to be the kind of phenomena that one has 

in mind when speaking of moving pictures in ordinary language’ (1996: 

131). I believe Carroll is in error here, for if he is attempting to establish an 

ontology of photographic film, then there should have been some reference 

to photography or film stock in his criteria. If he is seeking a broader 

ontology of moving images, however, then the ordinary usage of ‘motion 

picture’ for ‘traditional film’ is of little relevance. There is an inconsistency 

in investigating the moving image and restricting the results to its current 

popular incarnation. 

Inconsistency notwithstanding, there is a further problem with 

Carroll’s ontology in that it excludes both hand-crafted flip books (as he 

admits) and any moving picture presentation where the performance is an 

art. With regard to the latter, I have mentioned that the requirement that the 

moving image be generated by a template is an implicit link to Gaut’s 

mechanically-generated category. While the template might apply to mass-

produced thaumatropes and zoetropes – and perhaps even certain magic 

lantern shows – it would definitely exclude the shadow puppet plays, where 

the performance is not produced by a template and is itself a source of 

artistry. Perhaps this is not a fault in Carroll. If one considers watching 

Avatar in 3D and watching a shadow play, one might well conclude that 

they are not the same kind of thing. Yet Carroll’s ontology includes the flip 

book and the zoetrope, which seem much closer to the shadow play than the 

digital film in category. Why can a zoetrope be a moving image, but not a 

shadow play; and why can a mass-produced flip book be a moving image, 

but not a handmade one? Though Carroll’s ontology is remarkable in 

anticipating and including digital cinema, I believe it ultimately fails by 

excluding object-generated and handmade moving pictures without 

sufficient justification. 

Manovich’s characterisation of film as a particular type of animation 

may serve better, for animation seems to be – literally – the art of moving 

pictures. Manovich’s ontology of film is in fact more radical than this. He 

maintains that the return to handmade motion pictures epitomized in digital 

manipulation is not merely a return to animated moving pictures, but ‘a 

particular branch of painting – painting in time’ (1995). His answer to the 

ontological question is thus that a film is a series of paintings. The proposal 

is intriguing, but while there are aspects of digital cinema that are indeed a 

return to handmade motion pictures, the particular synthesis between human 

artistry and mechanical computer production in digital images appears to be 

a new development rather than a return to painting. One might, e.g., 

possibly conceive of Avatar as a series of paintings, but I’m not convinced 

that describing Rob Marshall’s Nine (2009) as series of paintings would be 

in any way enlightening. Although Guy Ritchie achieved the photorealism 

in Sherlock Holmes (2009) by animation, Manovich’s answer to the 

question ‘what is it’ still seems confusing. Even if one agrees with 

Manovich entirely, however, his ontology excludes traditional film, and the 

many similarities between digital and traditional film indicate that the 

former is a development of the latter rather than a new art form (in the same 

way that digital photography is a development of traditional photography). 

I believe that digital film is the same kind of ‘thing’ as traditional 

film and its predecessors, and that choreutoscopes and phenatakistocopes 

are similarly related to their own predecessors, magic lanterns and shadow 

puppet plays. They are all displays of motion pictures – or moving images 

– and while any two examples of the category may appear quite different, 

there is evidence of a linear development through history, particularly from 

the beginning of the nineteenth century. I agree with Manovich that the 

latest technology has asserted the dominance of animation over film, and 

reversed the relationship that persisted through the twentieth century. For 

all my criticism of Carroll, he had a seer’s vision in stating, ‘[t]he epoch of 

photographic film…may represent nothing but a brief interlude in the 

artform’ (1996: 122). Ultimately, however, Gaut’s description of the many 

kinds of cinema is the most convincing. Cinema, and moving pictures, have 

existed for centuries, and the current sovereignty of animation in digital 

imaging serves as a reminder that the moving picture pre-dated photography. 

What is a film? A series of pictures in motion. What is a digital film? The 

latest incarnation of the motion picture. Motion pictures can, as Gaut claims, 

be produced by objects, hand, or machine, but nonetheless belong to the 

same ontological category. 

The concept of film is currently in crisis, as it was eighty-odd years 

ago when the ‘talkie’ replaced the silent film. The talkie was seen by many 

as a threat, or a different kind of art form, but increased the opportunities for 

expression in the established art of film. The digital revolution has enriched 

the art form no less – perhaps even more so – returning the motion picture 

to its roots in animation, and creating new horizons for expression. Digital 

technology has created new possibilities for interactivity and mass-produced 

virtual reality. The consequences of virtual reality ‘total cinema’ and the 

threat of interactivity to the traditional unity of the art work are beyond the 

scope of my inquiry, and may produce questions to which a contemporary 

answer would be premature. At this point in the second decade of the 

twenty-first century, however, digital film remains – like traditional film 

and its predecessors – the art of moving pictures.