The Image as Direct Quotation: Identity, Transformation, and the Case for Fair Use

By Jaimie Baron

In contemporary digital culture, we are perpetually surrounded by a plethora of images whose barrage is in great and constant need of theorization if we are not simply to be swept away by it. To address this need, it is crucial that scholars be able to use and reuse the images that circulate around us in order to actively theorize them, their functions, and their effects. Moreover, in an era of online publishing in which images and video clips may be incorporated alongside text, the reuse of images and clips becomes an integral part of academic argumentation, enriching our understanding of visual culture by reframing and thereby re-comprehending fragments of its actual products. However, contemporary copyright laws and conventions frequently impede this vital practice by encouraging copyright holders to restrict the use of images and/or to seek licensing fees.

The U.S. Society for Cinema and Media Studies has been at the forefront of developing Best Practices in Fair Use for Media Studies scholarship in order to clarify how copyrighted material may be used in academic work. In addition to addressing questions about using films, clips, and film stills in teaching, SCMS has also attempted to articulate a clear set of guidelines for the use of copyrighted film images in academic publishing. Kristin Thompson has written a Report of the Ad Hoc Committee of the Society For Cinema Studies on the “Fair Usage Publication of Film Stills” in which she discusses at length the necessity for scholars to be able to include film stills as illustrations related to their scholarly arguments. She notes that, rather than using publicity or production stills, “For purposes of analyzing finished films…many scholars believe that photographs made from frames of the actual film strip are preferable, since they reproduce an actual composition that appears in a shot.” However, she notes, “The legal status of such reproductions of frames has remained problematic.” (1)

Thompson – writing for the Ad Hoc Committee – provides a very useful explanation of the cases in which the use of an image derived from a single film frame to illustrate a scholarly argument may most likely be considered an instance of fair use. Thompson’s analysis provides an excellent basis on which scholars and academic journals may justify their publication of film stills. (2) Indeed, Thompson makes a persuasive case for the film still as “transformative,” which is one basis upon which a claim of “fair use” may be made. Nevertheless, the threat that copyright holders may sue because they claim that their “property” has been “stolen” continues to inhibit the use of imagery in scholarship. In response, I would suggest that Thompson’s argument might be further strengthened – and expanded – by a phenomenological examination of the “reproduction” of an image and its recontextualization within a written text. Through an exploration of the “reuse” of an image as a form of “direct quotation” rather than simply a “reproduction” or “illustration” and as constituting a particular kind of viewer/reader experience, we may come to a more nuanced understanding of what may or may not constitute a “transformative” use of an image. Indeed, I argue that the reuse of an image – including but not limited to a film still – may, in fact, serve as a quotation and therefore may qualify as an instance of fair use.


The direct quotation as a form involves a tension maintained between identity and transformation. According to the 1871 book by Hiram Hadley entitled Lessons in Language: An Introduction to the Study of English Grammar, a direct quotation is “a repetition of the exact words of another.” (3) Given that Hadley’s book was published before many of the technologies of image production and reproduction were invented, it is no surprise that this definition refers only and specifically to “words” rather than to other kinds of signs. Even now, however, the “repetition” of an image (filmic or otherwise) is rarely, if ever, referred to in terms of quotation. Such repetition is much more frequently referred to as the “reproduction” of the image. However, I would argue there are reasons to consider certain repetitions of an image in general and a film image in particular within other texts as a form of quotation.

To begin with, in the case of both written quotation and image “reproduction,” in order to maintain the identity of the object as such, some things must be repeated “exactly” while others may be altered to some degree. When we quote a written source in a written text, what must be “exact” in its repetition is the identity and order of the words (or letters and spaces). In the case of written text quoting another written text, the “exact” form of the text does not extend to the font in which the original was printed, the size of the lettering, or even the layout of the letters on the page. What matters is that the letters and spaces retain the same relation to one another in terms of their order. The repetition of a written quotation in spoken form involves the transformation of one type of symbolic sign into another type of symbolic sign and from one medium to another; however, the spoken articulation is required to maintain only the order of words – emphasis (except where italics indicate emphasis in a written text), intonation, pitch, rhythm, and accent are not considered relevant to the identity of the quote. When written quotations are derived from spoken language, the quotation involves the opposite transformation of one type of symbolic sign into another, and from one medium to another. In this case, the question of whether to include “ums” and other sounds that may have interrupted the flow of words points to the difficulties that arise in quoting spoken language “exactly” in written language. However, in either case, a quotation of a previously written or spoken phrase is required to replicate only the precise identity and order of the words. It is important to note, however, that one may add, remove, or replace words from within the quotation through the use of ellipses and brackets. Although this convention is intended primarily to shorten a quote by omitting “irrelevant” words, removing and adding words nevertheless has the potential to disrupt and reconstitute meaning in a significant way. Thus, it is clear that the notion of the direct quotation as a “repetition of the exact words of another” is “exact” only in very particular ways. Phenomenologically – in other words, in terms of the experience of the reader – the quotation may have shifted in its design and layout (when writing is quoted in writing), its medium (when spoken language is quoted in writing or vice versa), and its continuity. Yet only the identity and order of the words matter to the identity of the direct quotation as such.

Like the direct quotation of a written passage, the “reproduction” of an image and the maintenance of its identity requires that some things be repeated “exactly.” Indeed, instead of the preservation of word identity and order, in the case of the image, what is crucial is that all visual elements retain the exact same spatial relation to one another. However, in order for these spatial relations to be exactly consistent, the image must necessarily be a mechanically copied representation with an indexical relation to the “original” (although the copy may sometimes be several times removed from the original). If the original image is not photographed – whether through analog or digital processes – but rather approximated by the human hand, the spatial relations may be altered (however slightly) which means that the image would cease to be an “exact” copy and would no longer function as “evidence” in the same way that an indexical copy produced by mechanical means would. Indeed, as numerous scholars have shown, the mechanically copied sign has historically been endowed with a privileged relation to the real and hence is more valued as “evidence.” (4) Thus, it appears that the image must be indexically, mechanically copied in order to function as evidence. (5)

However, just as the written quotation may be altered in some ways and still remain the “same” phrase, there are certain ways in which an image may be altered and still be considered the “same” image. For instance, the image may be altered in terms of its scale. This has generally been the norm in still photography. Analog photographs are generally derived from small negatives – the size of the negatives varying by format – from which a positive of (potentially) any size could be created. No matter the scale of the print, so long as it is based on the same negative, it is considered the “same” image. Digital photographs may also be resized both on the screen and when they are printed. 

This is clearly the convention in relation to the use of film stills in academic publishing as well. Indeed, the use of a film still within a written text is often referred to as a “frame enlargement.” According to the Society for American Archivists, a “frame enlargement” is “a photographic print made from a single frame of a motion picture.” (6) An example of a frame enlargement of Child of the Big City (Yevgeni Bauer, 1914) produced by the Harvard Film Archive and posted on the HFA Collections Blog reveals how a single image from a filmstrip is singled out for enlargement. (7)

The term itself suggests an element of alteration involved in producing a “frame enlargement,” but above the definition does not register this as an alteration, simply stating that the frame enlargement is “made” from a single frame. Moreover, the term also points to a phenomenological paradox. A frame “enlargement” is both larger and smaller than the image in the “film” itself. Assuming that it is mean to fit into a book or journal of some kind, it is necessarily larger than the tiny frame that makes up 1/24th of a second in a standard film reel. Were it not, it would be difficult for the viewer to see the image clearly. However, the “frame enlargement” is also much smaller than the projected cinematic image of which it is a constituent element – which, of course, may itself be projected at different scales. In terms of the viewer’s experience, however, the scale of the film image is, in fact, significantly reduced (unless, of course, the film was viewed on a very small screen or the film still is very large). However, as in the case of the still photograph, it is apparent that changes in scale are not generally regarded as constituting a significant transformation of the film image.

There appears to be an exception to this convention, however, when the change in scale is extreme. If the image is “reproduced” at such a minute scale that it ceases to be recognizable, it may not be regarded as the “same” image anymore. This seems to hold true in the case of small images collaged in order to create a larger image, such as the collage of thousands of tiny images from gay pornography arranged to generate a recognizable image of allegedly homophobic US Senator Rick Santorum’s face.

From the perspective of the viewer’s experience, the pornographic images are barely recognizable as such and appear to be “superseded” by the larger image of Santorum’s face. Other examples of photo mosaics are less ironic, for instance, an image of Barack Obama made up of newspaper covers from the day after his election, which reads more as an homage than a critique. (8

However, so far as I can tell, no one has sued over the use of tiny versions of their imagery in such “photo mosaics,” suggesting that no one considers these images as the “same” images now that they are miniscule parts of a larger whole. (9) Recognizability – a category which depends entirely on the viewer’s perception – seems to be the major factor in this convention. If the image is too small for the human eye to see clearly and it is part of a different image that the human eye can perceive clearly, it ceases to be the “same” image. (10)

Yet these conventions surrounding changes in the scale of the image are complicated by the fact that changes in scale are almost inevitably accompanied by an alteration in “quality” or “resolution.” As images are enlarged, they begin to lose their “definition” (a term whose multiple meanings are telling). This is certainly true in the case of the “frame enlargement.” Indeed, the Society for American Archivists definition above notes that “frame enlargements are generally of poorer quality than movie stills.” (11) In the case of an analog image, the larger the image, the “fuzzier” it will become as edges become less clearly defined. In the case of a digital image, the larger the image, the more pixelated the image becomes so that eventually it appears “blocky.” The question of whether a change in resolution can at some point produce a “different” image, however, has been a source of contention. This debate was crystalized in the lawsuit filed by photographer Jay Maisel, who shot the photograph of Miles Davis that was used on the cover of Davis’ album Kind of Blue, against Andy Baio, who produced a chiptune tribute to Kind of Blue entitled Kind of Bloop, generating synthesized music through the sound system of a vintage video game using Davis’ melodies. Baio licensed all the music; however, he also commissioned a pixel art recreation of the original cover – a photograph – but did not license the image since he believed that the pixel art recreation was sufficiently different from the photograph to be considered fair use. Even though, from the perspective of the viewer’s experience, the difference between the covers of Kind of Blue and Kind of Bloop constitutes a recognizable split between the experience of an indexical photographic image and that of a non-indexical (but, rather, loosely iconic) pixel image, Maisel sued and Baio ended up settling out of court. The cover is no longer available with the album although it can still be seen online. Baio’s blog, which relates this story in detail, also raises important questions about the relationship between resolution and transformation. He created a series of images, all derived from the Kind of Blue photograph with increasing degrees of pixelation, until the original image becomes unrecognizable. Next to this series of images appears the caption “Where would you draw the line?” (12)

Thus, the question remains whether there is some point at which a change in resolution produces a “different” image. If the degree of detail of the image is reduced so much that the elements of the image cease to “resolve” into a recognizable object, the viewer’s experience of the original image and of its unresolved “derivative” cease to have any coincidence. Nevertheless, it is clear that, according to current conventions (both academic and legal), relatively limited changes in the scale and resolution of an image (filmic or otherwise) are not usually regarded as constituting a “new” or “different” image. (13)

Another way in which an image may be changed yet continue to be regarded as the “same” image is that a color image may be reproduced in black and white. The “subtraction” of color, which makes “reproduction” less expensive, suggests that color is – at least in certain circumstances – not considered essential to the identity of the image. However, the reverse does not seem to be true. In the case of still photograph, Swedish artist Sanna Dullaway has been celebrated for the “artistry” involved in her colorization of iconic photographs of people such as Abraham Lincoln, Charles Darwin, and Anne Frank and of other historically significant photographs. (14)

John Hutchinson writes in the UK Daily Mail, that:

While purists may disagree with the modernisation of such historic pictures such as the celebrations that marked VJ (Victory in Japan) day, images of devastation from Pearl Harbour and the shocking shot of a Viet Cong prisoner being shot in the head by an army general (demonstrating the brutality of the Vietnam War), the colour pictures do appear to re-emphasise the scene they are portraying. (15)

This reference to “modernisation” and “reemphasis” suggests that Hutchison perceives Dullaway’s additions of color as producing a “different” image. To an even greater degree, it seems, color may not be added to a black and white film image and continue to be considered the “same” image, a fact to which the controversy of the late 1980s and early 1990s surrounding the colorization of black and white films attests. Scholars, filmmakers, and industry players argued for and against colorization on various grounds. (16) Interestingly, however, as Yuriko Saito has shown, one of the major objections to colorization was the feared alteration of the audience’s experience of the film. (17) Given the uproar surrounding colorization, it seems somewhat odd that the removal of color from a film image is considered unproblematic. From the perspective of the viewer’s experience, the transformation is just as significant. Nonetheless, “reproduction” of a color image in black and white is generally not considered an alteration of the image’s identity.

Thus, it appears that changes in scale and resolution (to some degree) and the subtraction (but not the addition) of color are not regarded as changing the identity of an image. In the case of the film still, however, there is another transformation involved: the transformation of a moving image into a still image. (18) Although film frames make up the underlying substrate of the film medium, they are never experienced by the film viewer as still images but rather as moving images, an experience that takes place over time. (This is slightly different in the case of the video frame, which we may experience as such when we pause the video. However, pausing the video is generally regarded as an interruption of the text rather than an integral part of the text.) If we redefine what is “the same” and what is “transformative” in terms of viewer experience, the film still and the film are identical for only 1/24th of a second (or a similar fraction for other formats with different frame rates), which is a period too brief for the human sensorium to identify as a distinct experience. Thompson notes that:

The commercial exploitation of films typically involves their being projected in such a way as to create the illusion of movement…Frame enlargements, however, do not duplicate the film in this way. A film frame, when printed on a page, cannot be projected as a portion of the original. It cannot create the illusion of movement, nor does it reproduce the sound that most films still in copyright involve. Even if we were to print every single frame of a film in a book, the result would in no way replicate the viewing experience. It is hard to imagine a person who has seen even thousands of frames reproduced deciding that he/she had “seen” the film and as a result had no need or desire to see it projected. (19)

Few would argue that still images and moving images belong to the same medium since, as Thompson clearly indicates, the phenomenological disparity between the viewer’s experience of the film still and of the moving image is so great. To argue that the film still is a “reproduction” of the film is to assume that reproduction refers only to a replicated object (the frame of film) rather than to a reproduced experience. And yet, when we remember a film we have seen, we remember the experience of seeing moving images, not the series of still images on a strip of wound celluloid sitting in a can. Hence, the film still cannot be reduced to the “same” experience as watching a film even if it is produced from the film object. Nonetheless, the film still is generally regarded as a “reproduction” of a film at least insofar as copyright holders continue to seek licensing fees.

Thus, we can see that in the case of both written quotation and image “reproduction,” certain elements (though not precisely the same ones) may be altered while the identity of the phrase or image is considered to remain fundamentally “the same.” However, because of the requirement that an image be an indexical, mechanical reproduction of the original in order to remain the “same” image, there is no middle ground upon which an image may be directly “quoted” without also being “reproduced.” This creates a double bind for those who wish to “reuse” an image. An approximation of the image is not “direct” enough to act as significant evidence while an indexical, mechanical reproduction of the image is too “direct” and raises the specter of copyright infringement. The “permissible” changes in scale, resolution, color, and motion do not alter the identity of the image precisely because they do not “count” as transformative. (This despite the fact that, from the perspective of the viewer’s experience, changes in scale, resolution, color and motion – at least changes of a certain degree – are potentially very transformative.) Indexical, mechanical replication is the only thing that can both insure that the image is the “same” while simultaneously opening anyone who uses the “reproduced” image up to charges of copyright violation. I would argue that we should reframe indexical, mechanical reproduction – in other words, the fact that all of the elements of the image retain their exact spatial relation – as parallel to the requirement that the words in the direct written quotation retain their “exact” order. Spatial relations in the image “reproduction,” like word identity and order in the direct written quotation, becomes the ground upon which the identity of the image is established and against which the new writer may construct new ideas.


Indeed, the value of a direct written quotation does not lie simply in its identity as such but also in its “productivity” within its new context, this productivity made no less important by its dependence on a subjective evaluation on the part of the reader. While the identity of the direct written quotation must be maintained, its significance must be transformed through its recontextualization, producing something “new.” Indeed, although it is crucial that certain things remain the same within the quotation marks in a written quotation, it is also imperative that something “different” be produced in the relation between what is inside and what is outside the quotation marks. Of course, any repetition in a new context alters the quotation and produces something “new” and “different.” As Meir Sternberg notes of written quotations:

The framing of an element within a text entails a communicative subordination of the part to the whole that encloses it. However accurate the wording of the quotation and however pure the quoter’s motives, tearing a piece of discourse from its original habitat and recontextualizing it with a new network of relations cannot but interfere with its effect…To quote is to mediate and to mediate is to interfere. (20)

Indeed, the act of recontextualizations always “mediates” or “interferes” with the “effect” of the quoted material. And yet, if “too much” of the original is quoted, it may seem that the new text is simply showcasing the original text rather than effectively reframing and thereby transforming its meaning in some way. Moreover, if the written quotation is not fully “unpacked” in relation to the new text in which it appears, the “interference” may not be considered a “productive” use of the quotation. This notion of a “productive” use seems to occur in the interaction between that which is inside the quotation marks and that which is outside – as if there were a chemical reaction that might or might not take place. If it does not occur, the quote is “wasted.” In addition, when a quotation is imported into a new text, to be productive, it seems that the quote must be fully “integrated” into the new text. To insert the quote without integrating it into a new text, would simply be to put it in a new context – to just repeat it – rather than to make it part of a new text – to repeat and, in doing so, transform it. Indeed, it is the tension between something “exactly” the same within the quotation marks and something different produced between what is inside and outside the quotation marks that defines the ontology and productivity of the written quotation. 

The problem of taking “too much” differs in the case of the still image and the film still. In the case of the still image, it is impossible not to take “too much” while maintaining the identity of the image. Taking a part of a still photograph, for instance, (except as a detail of a larger image, also presented in full) would alter the spatial relations of the image. Even cropping the image is generally considered an alteration of the identity of the image. As a result, as I have established above, if only part of the image is taken, it is no longer the same image. In the case of the film still, the tiny fraction of the film (and filmic experience) it embodies suggests that it cannot be regarded as taking “too much” of the film itself. However, a film still reads as a “complete” image, which may be why it continues to raise copyright issues. Despite these different relations between still images and film stills and their source texts, I would nevertheless argue that the still image can be transformed by its recontextualization in the same way as the film still can.

Indeed, it follows from the discussion above that the productivity of the image in a written essay hinges on a transformation in relation to what is immediately external to the “frame” that marks the boundary between the image and what is outside the image. The image must interact with its new environment to produce something “new.” However, while the written quotation can be “integrated” into the new written text by virtue of being of the same medium, this is not true for the image “quoted” within the written text. Indeed, because still images are not of the same medium as writing, they are usually regarded as “illustrating” the written text rather than being fully “integrated” into it. 

This distinction between “integration” and “illustration” is crucial in that it is, in part, the written quotation’s successful integration into the new written text that makes it seem “productive.” The written quotation is marked as “other” but it becomes a piece of the written text into which it is inserted. It does not illustrate but rather contributes – almost but not quite seamlessly – to the text at hand. Experientially, although the reader’s eye will note the presence of quotation marks, the eye may continue reading the quotation without a pronounced spatial or temporal interruption. A written quotation of a written or spoken text remains similar to its source in that the source, like the quote, is experienced over a period of time. Thus, the quoted text and the new text can be experienced as an uninterrupted flow.

An image, however, cannot be “integrated” into a written text in the same way. When we are reading a written text, if an image accompanies the written text, our eyes must stop reading the words in order to take in the nearby image. Although we may look at the image for a long or short time, this act of looking is experienced (if unconsciously so) as an interruption. We must stop reading, look at the image, then look back at the written text. As a result of this disruptive experience, the image remains more noticeably “other” to the written text than does the written quotation inside quotation marks. I would argue that it is, at least in part, this shift in the mode of “reading” that leads us to refer to an image within a written essay as an “illustration” rather than a “quotation.” (21) Yet, in an academic essay about an image, if the written text is analyzing the image, it seems disingenuous to argue that image simply “illustrates” the analysis. If we regard the frame of the image (the boundary between inside and outside the image) as similar to a set of quotation marks, it is the interaction between inside and outside that has the potential to produce a transformative effect and a new significance. Yet we have a very limited vocabulary with which to discuss this interaction between image and written word, so the interaction is reduced to “illustration,” a term that implies the image to be subservient to the written word, an auxiliary rather than integral and productive part of the new text. (22) It is worth noting that a similar problem occurs in the case of online journals that include film clips along with the written text. Although the film clip, like the written text, is experienced in time, the switch from reading to viewing a clip is also likely to be experienced as an interruption. The film clip cannot be seamlessly integrated into the written text any more than the film still can.

Despite the fact that an image cannot be fully “integrated” into a written text, I would nevertheless argue that we need to reconceive of the interaction the words of a written essay and the accompanying image as not an illustration but rather as a quotation because it produces – or has the potential to produce – something “new.” While the experience of a written text and an image cannot be seamlessly united, the inclusion of an image may allow for a productive interaction between written text and image. Indeed, when an image appears with a written text about that image, my experience of seeing the image under discussion is often one of epiphany. Something that was abstract and unspecific when it was described in only written words becomes concrete and particular as I gaze at the image. I am able to reexamine the image within the context of the argument being made and, as a result, new readings of the image become possible. Thus, despite the lack of seamless integration involved in using an image in a written text, a “something new” is generated between the image and its new context. By the same token, however, if the image is not fully “unpacked” in the text, it may not constitute a “productive” reuse of the image. In the case of a film still, if the still is included simply as a visual metonym for the film and is not directly discussed, then it may retain its status as an “illustration” rather than as a productive quotation. If they are to be considered quotations, images must not simply serve as “decoration” for the written text. Words and image must collide and in their collision transform their respective significances and, in doing so, also transform the viewer’s experience of each one. This is, of course, the basis of a claim of fair use.

Implications for Remix Film and Video

While the argument above is focused on the use of images in academic publishing, the discussion of the “quotation” of an image also has important implications for thinking about the use of film clips in “remix,” a broad term for films and videos that appropriate preexisting audiovisual materials. While, as established above, an image cannot be fully integrated into a written text because it exists in another medium which our sensoria experience differently, using a film still or a film clip in a film or video would seem to achieve a greater parallelism between the quotation of an image and a quotation of written phrase. When filmmakers appropriate stills or clips – audio, visual, or both – into a new text, this audiovisual quotation may be integrated into the new audiovisual text just as a written phrase may be integrated into another written text with the minor intervention of quotation marks. Of course, just as a written quotation must be marked as such by quotation marks, it is arguable that filmmakers should also mark their appropriations by citing their sources in some form or another. Although a standard format for citing visual sources in a film has yet to be developed, as theorist Patrik Sjöberg has noted, there is no reason that this cannot be done. (23)

Those who pursue charges of copyright infringement, however, do not profess to care about citation. Rather, they simply assume that visual media cannot be “quoted” since the term “quotation” is associated only with words. By this logic, the mechanical reproduction of any portion of an image, still or moving, is an act of theft. However, if we reimagine the mechanical reproduction of audiovisual clips as a form of “direct quotation,” whose identity must be maintained as such in the same manner as the identity of the direct written quotation, and if we require that the incorporation of this audiovisual “quotation” into a new audiovisual text transform the significance of both the clip and the surrounding text, then such audiovisual appropriation is not an act of theft so much as one, as Sternberg suggests, of productive “interference.” Indeed, academic scholarship, artistic creation, and the production of new knowledge itself are based on the act of “interference.” The role of the intellectual or artist is precisely to “interfere” with sedimented ideology that masquerades as and is blindly accepted as “universal truth” or “common sense,” preventing us from imagining other, potentially better ways of thinking and being. If we cannot “interfere” with audiovisual texts through “quotation” in the same way that we regularly “interfere” with written texts, then the academic and artistic projects may come to a halt. I would argue that in a society in which communication increasingly takes place through audiovisual rather than written forms, we must begin to rethink quotation as a multimedia tool. 

Moreover, the above comparison between the direct written quotation and the film still as a form of direct quotation also suggests that we must rethink fair use less in terms of a transformed object than a transformed viewer experience. Of course, as I have demonstrated, certain slight transformations of viewer experience of the object such as relatively minor transformations of scale and resolution and the removal (but apparently not the addition) of color, are generally regarded as, by and large, insignificant to its identity (although many film purists would argue that changes in quality and resolution make a major difference in terms of viewer experience). Major transformations of scale and resolution, however, do change viewer experience to a degree that the object (as co-constituted by the text and the viewer) can no longer be regarded as the “same.” Moreover, the shift from a time-based to a non-time based form in the case of the film still or the removal of a short temporal segment from a longer piece in the case of the film clip, also significantly alters viewer experience. A film still and a relatively brief film clip do not provide the same experience as the complete film text. Where we draw the line between a significant and insignificant change in viewer experience will continue to be a point of contention. However, I believe that rethinking the reuse of still images, film stills, and film clips within academic or artistic texts as producing an experience parallel to that of reading a direct written quotation within another written text may begin to clarify this distinction. And where the repetition of an image or clip functions as or like a direct quotation, it must be considered an instance of fair use.


(1) Kristin Thompson, “Report of the Ad Hoc Committee of the Society For Cinema Studies, ‘Fair Usage Publication of Film Stills,’” Accessed 9 April 2012. See Thompson’s essay in this issue of Frames in which she refers to her work in this area, especially in relation to online publishing, and also Steve Anderson’s essay, which examines Fair Use arguments.

(2) I use the term “film still” to refer to single film or video frames, not to be confused with the terms “publicity still” and “production still,” which generally refer to a still image taken on the set of a film and intended primarily for publicity purposes. I use the term “film image” to refer to the source of the film still.

(3) Hiram Hadley, Lessons in Language: An Introduction to the Study of English Grammar (Chicago: Hadley Brothers, 1871), 16.

(4) See, for instance, Bill Nichols, Representing Reality: Issues and Concepts in Documentary (Bloomington: Indiana University Press, 1991), 149-155.

(5) This distinction between indexical and non-indexical copies is complicated, however, by extremely realistic hand-drawn copies of photographs such as those drawn by Scottish artist Paul Cadden. His drawings are so similar to photographs that they appear indexical even though they are not. See Huffington Post, “Paul Cadden’s Unbelievably Photorealistic Drawings,” 26 April 2012. Accessed 14 May 2012.

(6) Society of American Archivists, Accessed 12 April 2012.

(7) Harvard Film Archive Collections Blog, 9 February 2009. Accessed 14 May 2012.

(8) Kevin Hoffman, “Five best Obama Photomosaics on Flickr,” 22 December 2008. Accessed 14 May 2012.

(9) Interestingly, however, artist and programmer Robert Silvers has trademarked the term “Photomosaic” and patented his Photomosaic software which allows users to easily produce such images. See his website at Accessed 8 May 2012.

(10) Another reason there have been no lawsuits about specific photo mosaics may be that many such photo mosaics are posted online and are not intended for sale or other commercial purposes.

(11) Society of American Archivists. , Accessed 12 April 2012. This usage of “movie still” appears to refer to publicity stills.

(12) Andy Baio, “Kind of Screwed,” Waxy, 19 July 2011. Accessed 12 April 2012.

(13) Similar questions were at stake in the better-known case of Shepard Fairey’s Obama Hope poster, which he derived from a photograph taken by AP photographer Mannie Garcia. Fairey took Garcia’s entire image but recast it in large blocks of color so that it ceased to look like an indexical photograph and more like an iconic painting of Barack Obama. Although Fairey claimed his use of Garcia’s image to be “transformative” and therefore “fair use,” like Baio, he ended up settling out of court. For a useful discussion of the Fairey case that begins to take into account audience experience, see H. Brian Holland, “Social Semiotics in the Fair Use Analysis,” Harvard Journal of Law & Technology 24, no. 2 (Spring 2011): 335-391.

(14) See for instance, Pratik Naik, “Unbelievable colorization of black and white images,” Accessed 8 May 2012.

(15) John Hutchison, “The past just got a whole lot more colourful: Painstaking touch-up of world’s most famous black and white pictures,” Mail Online, 19 January 2012. Accessed 8 May 2012.

(16) For extended discussions of film colorization, see Michael Dempsey, “Colorization,” Film Quarterly 40, no. 2 (Winter, 1986-1987): 2-3; Flo Leibowitz, “Movie Colorization and the Expression of Mood,” 
The Journal of Aesthetics and Art Criticism 49, no. 4 (Autumn, 1991): 363- 365; James O. Young, “Still More in Defense of Colorization,”
 The Journal of Aesthetics and Art Criticism 50, no. 3 (Summer, 1992), 245- 248; Yuriko Saito, “Contemporary Aesthetic Issue: The Colorization Controversy,” Journal of Aesthetic Education 23, no. 2 (Summer, 1989): 21-31.

(17) Yuriko Saito, “Contemporary Aesthetic Issue: The Colorization Controversy,” Journal of Aesthetic Education 23, no. 2 (Summer, 1989): 23-24.

(18) I use the term “film still” in part because it emphasizes the stillness of the single frame in contrast to the moving image. Other synonymous terms such as “frame capture” or “frame grab” evoke the sense of an object violently excised from another context and from a different order of experience.

(19) Thompson, Accessed 9 April 2012.

(20) Meir Sternberg, “Proteus in Quotation-Land: Mimesis and the Forms of Reported Discourse,” Poetics Today 3, no. 2 (Spring 1982): 108.

(21) Of course, an image can be “integrated” into some form of collage or other visual text, a point to which I will return later.

(22) Scott McCloud’s Understanding Comics, 2d ed. (New York, Harper Collins, 1993) might provide an interesting way of thinking about the relationship between written text and image in academic film writing.

(23) Patrik Sjöberg, The World in Pieces: A Study of Compilation Film (Stockholm: Patrik Sjöberg, 2001), 46.

Works Cited

Baio, Andy, “Kind of Screwed.” Waxy, 19 July 2011. Accessed 12 April 2012.

Dempsey, Michael, “Colorization.” Film Quarterly 40, no. 2. (Winter, 1986-1987): 2-3.

Harvard Film Archive Collections Blog, 9 February 2009. Accessed 14 May 2012.

Hoffman, Kevin, “Five best Obama Photomosaics on Flickr.” 22 December 2008. Accessed 14 May 2012.

Holland, H. Brian, “Social Semiotics in the Fair Use Analysis.” Harvard Journal of Law & Technology 24, no. 2. (Spring 2011): 335-391.

Huffington Post, “Paul Cadden’s Unbelievably Photorealistic Drawings.” 26 April 2012. Accessed 14 May 2012.

Hutchison, John, “The past just got a whole lot more colourful: Painstaking touch-up of world’s most famous black and white pictures.” Mail Online, 19 January 2012. Accessed 8 May 2012.

Leibowitz, Flo, “Movie Colorization and the Expression of Mood.”The Journal of Aesthetics and Art Criticism 49, no. 4 .(Autumn, 1991): 363- 365.

McCloud, Scott, Understanding Comics, (2d ed. New York, Harper Collins, 1993).

Naik, Pratik, “Unbelievable colorization of black and white images,” Accessed 8 May 2012.

Nichols, Bill, Representing Reality: Issues and Concepts in Documentary. (Bloomington: Indiana University Press, 1991).

Saito, Yuriko, “Contemporary Aesthetic Issue: The Colorization Controversy.” Journal of Aesthetic Education 23, no. 2 (Summer, 1989): 23-24.

Silvers, Robert, Accessed 8 May 2012.

Sjöberg, Patrik, The World in Pieces: A Study of Compilation Film. (Stockholm: Patrik Sjöberg, 2001).

Society of American Archivists, “Frame Enlargement.” Accessed 12 April 2012.

Sternberg, Meir, “Proteus in Quotation-Land: Mimesis and the Forms of Reported Discourse,” Poetics Today 3, no. 2. (Spring 1982): 107-156.

Thompson, Kristin, “Report of the Ad Hoc Committee of the Society For Cinema Studies, ‘Fair Usage Publication of Film Stills.’” Accessed 9 April 2012.

Unicorn Booty. Accessed 5 May 2012.

Young, James O. “Still More in Defense of Colorization.”The Journal of Aesthetics and Art Criticism 50, no. 3 .(Summer, 1992), 245- 248.


Frames # 1 Film and Moving Image Studies Re-Born Digital? 2012-07-02, this article © Jaimie Baron. This article has been blind peer-reviewed.