Print | PDF
Amodern 8: Translation–Machination
January 2018


Archive-Based Image Description as an Intermedial Translation Technique

Jane Birkin

I admit to a longstanding preoccupation with image description, the primary focus of my theoretical and artistic research, and a peripheral knowledge of the science of interlingual translation. Nevertheless, the connections between description and translation can be explored within the specific context of intermedial translation techniques and from the vantage point of image description. The description of the visual content of a single photographic image, with its foundations in archival cataloguing practices, operates across the technical and linguistic boundaries of the media forms of text and image, and there are relationships between the activities of image description and interlingual translation that can be fruitfully mapped and probed. Translation scientist Wolfram Wilss argues that translation is “a special case of meaning-based information processing.”1 When applied institutionally, description is a form of information processing and information management and can be compared in this respect to literal translation. Literal translation is perceived as “the direct transfer of an SL [source language] text into a grammatically and idiomatically appropriate TL [target language] text, a translation between two languages of the same family and sharing the same culture.”2 Here the photograph and the description are defined as parallel media, sharing the same informational and temporal characteristics and performing common roles in the organization and processing of cultural knowledge.

Two translation experts whose books I make frequent reference to in this essay approach translation from different corners, although they frequently overlap. Lawrence Venuti, in his edited volume, The Translation Studies Reader (2000), presents a theoretical view of translation studies as a (then) emerging academic field, whereas Wilss, in Knowledge and Skills in Translation Behaviour (1996) puts forward a primarily pedagogical and practically rooted rationalization of translation as a skill, albeit supported by translation studies. Wilss terms the process of giving an explanation of how a text is translated as “metatranslational.”3 He explains: “The difference between translation and metatranslational acts corresponds to the distinction between practice-oriented and theory-oriented (epistemic) knowledge.” This essay makes a strong argument for the raised status and even the autonomy of description and, in addition to providing various examples in support of this position, it foregrounds a reproduction of my own set-piece description of a found news image, Patrons–watch–an–activist–004.jpg.4 This image description, written in 2014, serves not only as a direct and indirect reference point for theoretical arguments made, but also as an example of practice-oriented knowledge production, that is, within a practice-as-research art milieu. The rest of the essay could be classed as “metatranslational” dialogue that may enhance the understanding of the process of description, but not the description itself. Bruno Latour5 identifies a perception that description is risky and he argues:

[W]e worry that by sticking to description there may be something missing, since we have not “added to it” something else that is often called an “explanation.” And yet the opposition between description and explanation is another of these false dichotomies that should be put to rest […] if a description remains in need of an explanation, it means that it is a bad description.6

I hope that Patrons is a good description. At any rate, it can be approached as a discrete text and read without explanation.


Outside the Archive

Patrons–watch–an–activist–004.jpg is an example of an extended content-based description of a journalistic image.


The description defines its parameters at the start, so as to afford some basic understanding of what is to follow: “This is a photograph of the interior of a cafeteria or restaurant, viewed from the outside.” It is identified as a photograph that exists or, at any rate, existed at the time of its description, yet it remains unseen here, and this is to my advantage: readers of the Patrons description are oblivious to any difference between image and description. As Venuti argues of the translated text: “The loss in translation remains invisible to any reader who does not undertake a careful comparison to the source text – i.e., most of us.”7 And Venuti emphasises his first rule of reading translations: “Don’t read just for meaning, but for language too; appreciate the formal features of the translation.”8 The restricted, institutional language used in the Patrons description carries forward from the archive a form and a formality. The direct language of image description is part of a controlled information management system, but in the case of Patrons it is also an applied methodological construct and it is taken to an extreme: it follows the rules of archival object-level description, but the length and the detail of the description are only possible because it is written outside of the time constraints of archival labour. This set-piece description is positioned as a piece of conceptual writing, as an artwork, where the making is accepted as time intensive.

Inside the archive, hierarchical levels of description present a way in to the understanding of context, duration or progression of time, as they travel from “the general to the specific,” beginning with the description of the fonds, that is, the entire body of records of an individual or organization, and ending in the description of the single object.9 The only contextual markers offered here for Patrons are the headline, credit and source: metadata that materialized with this digital news image. There are no supporting archive materials, no sense of “part-to-whole relationship” to add context, duration or temporal progression.10 This discrete description connects us back to the time of the apparatus, as it corresponds to the camera shutter as an instrument of capture. Classicist D.P. Fowler, in his essay “Narrate and Describe,” remarks on the correspondence between camera and language in a description by novelist Leonardo Sciascia (from his 1987 work 1912+1): “Sciascia attempts to describe the scene neutrally, like a camera with the shutter open.”11 There is direct reference made in the Patrons description to the camera and its connection to the embodied time of the image: “His arm and hand are blurred, probably due to movement and the slow shutter speed of the camera in the limited available light.” And, without reference to the camera, the instant of capture is translated by the description of the momentary and time-critical configuration of the two men’s arms: “This isolated arm is raised in a similar fashion and at a similar angle to the first man’s, so that the two arms configure as parallel forms.” The two arms are held in stasis, preserved for time to come, a direct analogy to the temporal complexities of the archive itself.

Kari Kraus argues that in Charles Henry Middleton’s nineteenth century catalogue description of Rembrandt’s famous etching of Dr. Faustus, the “enumeration of the desk, chair, books, hourglass, and skull,” means that it is “likely to strike a reader who has grown up in a cultural milieu saturated by images […] as peculiar in its single focus on object identification.”12 The Middleton description nevertheless lies within the bounds of cultural processing and can be understood as such; the transfer of the archival description technique to a different cultural space (appearing as an artwork, or as an artwork that is part of a publication) is transformative. The appropriation of institutional techniques of image description openly contests traditional modes of image critique and the language of description appears radical when placed in these spaces. Fowler goes further, stating that “the more radical move is to free description from the chains of slavery and to give it true autonomy.”13 Such language of slavery and liberation is commonplace in dialogue on the archive image. For example, Hito Steyerl describes the “poor image” as “liberated from the vaults of cinemas and archives and thrust into digital uncertainty.”14 The idea of liberation is tied up with notions of the readymade and the dispositif, the transformation of an object through its positioning, and image appropriation and associated tactics of reordering and re-contextualization have become mainstream artistic strategies, especially in the age of the networked image and with artists who work with archive media. The traditional institutional role of the description is to stand in place of the image, becoming redundant once the image is viewed. In contrast, I am arguing for their parity. The description must enjoy equal freedoms, and be experienced as a transformed object, outside of its natural milieu.


Object-level and Content-based

Walter Benjamin asks: “Is translation meant for those who do not understand the original? […] it seems to be the only conceivable reason for saying “the same thing” repeatedly.”15 In both translation and description the reception of the new text relies on the reader’s confidence in equivalence to the source object, which is concealed or obscured: the purpose of image description in the archive is not to support the understanding of a visible image, but to impartially represent the hidden image that it catalogues and describes. To be effective, it must say the same thing. Kraus identifies two valid modes of comparison in picture criticism: “collocation”, where the two objects are placed side by side for comparison, and “collation,” where they are separate. She links this to technical developments that have allowed, over time, reproductions of images to be placed in catalogues alongside their descriptions.16 She explains:

[W]hat becomes clear is that the relationship between collation and collocation fundamentally shifts over time. In the nineteenth century — prior to the rise of photomechanical reproduction — the descriptive catalog was generally devoid of images, and consequently collocation played little role in the presentation and analysis of variants. Instead, description and collation combined to serve a substitutive function, standing in for the missing visual objects.

Technically speaking, the object-level image description of visual content originated before images were easily reproduced and shared (digitally or otherwise). Archival descriptions of hidden images could thus be regarded as past, even “dead,” media forms, except that extant descriptions are often left in place and descriptions continue to be written when resources allow. In media-archaeological terms, object-level description is one of many archive cataloguing and organisational techniques that influence and question network call-up and storage behaviours. The word “archive,” furthermore, has itself been incorporated into our digital vocabulary. Yet the archive and the network enjoy a somewhat problematic relationship: the network, with its “chaotically shelved” and precarious content, sits in opposition to the notions of custodial care and stasis that are at the core of the archive project.17 Meanwhile, vast numbers of archive photographs are not accessible as images online, but are stored in archive boxes in locked strongrooms. They are often first encountered through text descriptions in paper catalogues or stand-alone databases. The image description is written in prose and is a catalogue of the visual content of the image. Contextual information is usually stored at higher levels of description (collection- or folder-level) leaving the object-level description predominately context free, offering an unusual way of writing and reading images. Description of the single object, defined by the General International Standard Archival Description as the “smallest intellectually indivisible archival unit,” is in many institutions reserved for collections deemed the most historically or culturally significant.18

Object-level descriptions are more than mere anachronisms. In addition to continuing as a valuable offline resource, they can successfully migrate into searchable texts, positively existing outside of limited vocabularies and predefined thesaurus terms. Financial constraints mean that the use of specialised metadata schemas to add keywords to images is increasing, as workers with little or no professional training can carry this out cost effectively. Free labour is used as images are put out for public tagging by large institutions such as the Library of Congress.19 Although the use of common schemas increases operability between archive collections, metadata schemas actively encourage the recording of relationships between items, and not the items themselves. Often operating above object-level, they describe common events, not discrete situations, the general, not the specific, and frequently result in skewed search returns. There is a fundamental breach between image and keyword. The single word descriptor takes on a different operative relationship to the image than does full image description, which builds a static and singular connection, one to the other, exposing what is actually visible, rather than speculating about what might be implied. Stephen Connor (2002, n.pag.) highlighted the restrictions of searching within a limited vocabulary in his talk to the Friends of the University of London Library:

the keyword search can easily shrink into a kind of keyhole surgery, in which what you get out is too narrowly prescribed by what you put in. […] You need to know not only the kinds of words that others have used to designate your topic, but the lexical and intellectual company those words have kept.20

Although objects and images are indeed connected across archives and institutions by the same keywords, these words may struggle as shared descriptors, rendering interoperability somewhat superficial.

Translation theorists and professors Jean-Paul Vinay and Jean Darbelnet see the possibility of literal translation primarily as a way to translate technical and scientific texts: “largely based on the existence of parallel passages in SL and TL texts, corresponding to parallel thought processes which, as would be expected, are particularly frequent in the documentation required in science and technology.”21 The translation of the technical text, I contend, parallels the description of the technical image (the photograph) in the archive: they are both content-based informational objects, requiring no interpretation or addition of meaning by the archivist/translator. Outside of basic technical texts, Vinay and Darbelnet argue that literal translation can only go so far, until an obstacle is encountered. The translator must then focus on the message and the “totality of the message” drives the process. They regard literal translation as limited, as undermining the broader theoretical underpinnings of translation, its unambiguous transfer from source to target language perceived as lacking intellectual challenge.

Description has similarly been seen as a lower form of writing. In a recent special issue of Representations, which examines description across a range of disciplines, Sharon Marcus et al strongly urge us “to stop taking the lesser status of description for granted.”22 D.P. Fowler had argued against the disparagement of the form 25 years earlier,23 by citing Nouveau Romancier Alain Robbe-Grillet, who stated of description that “instead of this universe of “significations” (psychological, social, functional), one must try to construct a world more solid, more immediate.24 Nouveaux Romanciers such as Robbe-Grillet and Georges Perec used descriptive techniques to afford an equality of denotation that meant that nothing much was signified at all. In Perec’s book Things: A Story of the Sixties, the objects in the lives of the materially motivated couple, Jérôme and Sylvie, are described scrupulously; the characters are viewed through their things, their possessions and surroundings, both existing and desired. Their many desires are marked out not by dialogue or narrative, but by Perec’s use of the conditional tense:

On each side of the table, virtually facing each other, would be two high-backed wood and leather armchairs. Still further to the left, along the wall, would be a narrow table overflowing with books. A wing-chair in bottle-green leather would lead to grey metal filing cabinets and light wooden card-index boxes. On a third, even smaller table would be a Swedish Lamp and a typewriter under its canvas dust-cover. Right at the back would be a narrow bed covered in ultramarine velvet and stacked with cushions of all colours.25

Perec’s repetitive listing of material objects and situations render them valueless and the couple’s desires appear absurd.


Creativity, Intuition and Shared Knowledge

In translation studies, the connected theories of “translation creativity” and “translation and intuition” allow translators to work creatively and intuitively within the confines of a given text, whilst preserving the original message. Benjamin, with respect to creativity, though agreeing that the message (“intention”) is critical, sees translation as a technique quite different from other forms of writing and observes that a “great poet” will not necessarily produce a better translation than a lesser one.26 He explicates: “As translation is a mode of its own, the task of the translator, too, may be regarded as distinct and clearly differentiated from the task of the poet.” Wilss goes further, arguing that creativity might be contradictory to the nature of translation, whose goal is, put simply, “to reproduce ST [source text] in a TL.”27 He thinks that more open and creative methods can only be used by experienced operators: “Translation creativity is some trait that can be expected of a translator who has accumulated a wide range of translation knowledge and can now apply this knowledge appropriately and judiciously in translation circumstances.” Wilss sees translation intuition as closely linked to translation creativity, and similarly difficult to formulate procedural rules for:

All translators will orient themselves, whenever possible, toward the procedural patterns that they have acquired in a more or less systematic way, and they will tend to produce a methodologically and linguistically institutionalized form of language usage. But they must always be prepared for situations that lie beyond the standardized modes of translation. This is where translation intuition comes in.28

He warns of the dangers and unpredictability of intuition; it should perhaps only be used in emergency situations, and there is no guarantee of success.29 In a similar vein, there may be situations with description that lie beyond standardised modes, things that cannot be confidently described because they are not clear from the visible image. Within description, however, creativity and intuition are kept in check, and any speculation is recorded in the text itself. The description writer, that is, presents ambiguous situations in guarded terms with such phrases as “there is a possibility” or “there appears to be.” The writer is reporting on the possibility of something: a possibility that is contained in the image and transmitted through the description, in effect circumventing the binary responses of metadata schemas and search engines.

Explicit knowledge is the primary currency of archival description, but some tacit knowledge must be assumed. For example, one must accept common knowledge of the existence of objects such as doors, floors, ceilings, hats, wedding rings, chairs, tables and newspapers, as well as their general significance in the world. The notion of “schemata” in translation theory facilitates communication and comprehension by taking into account shared prior knowledge and related cognitive and linguistic structures, as well as human behavior patterns. Federica Scarpa explains that it refers to

abstract structures representing chunks of knowledge about the world (events, actions, situations) […] which deal with knowledge about the properties of objects (e.g. the idea of KNIFE) and locations, and “scripts,” which deal with knowledge about events and sequences of events (e.g. a visit to the dentist).30

In description, not to assume this kind of basic knowledge would result in the rendering of the description into a reductive notation of shape and colour. Intriguingly, this is exactly what Kenneth Goldsmith is endeavoring to do in describing Ellsworth Kelly’s 1951 abstract painting “Seine” in what he calls “insanely precise detail.” It is a lengthy description project, with an extremely limited vocabulary:

Several parallel units, comprised of left aligned vertical elements dominate this side of the painting. Twenty-eight and Thirty are one and an eighth inches from the left edge; Twenty-seven, Twenty-six, and Seven are one and five-eighths inches from the left edge; Eight, Ten, Five and Four are two and three-sixteenths inches from the left edge; Twenty-nine, Nine, and Three are two and a quarter inches from the left edge; Twenty-four, and Two are three and a quarter inches from the left edge; Thirty-one, Thirty-four, Eleven, Thirteen and One are three and three-quarters inches from the left edge;31

In this case, the image itself is made up of reductive notations of shape and colour. These limitations (restricted to index numbers, measurements, colours, locations) directly match the limitations of the visual elements, making Goldsmith’s piece a conceptually sound example of direct image to text translation.

Erwin Panofsky, in his text on iconology, cites three strata of subject matter or meaning contained in an image: primary, where the viewer must have a “familiarity with objects and events,” secondary or conventional, where a “familiarity with specific themes and concepts” is needed, and intrinsic, where one would require a “familiarity with the essential tendencies of the human mind.”32 Correspondingly, Roland Barthes defines his three levels of meaning: firstly, the informational, or communication, “which gathers together everything I can learn from the setting,” secondly, the symbolic, signification, or the obvious, “a second or neo-semiotics, open no longer to the science of the message but to the sciences of the symbol,” and thirdly, significance, or the obtuse, a meaning which extends “outside culture, knowledge, information.”33 Visual, content-based description deals almost exclusively with the primary and informational. It could stray into the secondary, but stops short of Barthes’ symbolic and the third levels of Panofsky’s intrinsic and Barthes’ significance. Kraus, continuing her discussion on the hidden and the visible image, places Middleton’s description of Rembrandt’s Dr. Faustus within Panofsky’s three levels at the first (pre-iconographic) level, “a literal interpretation of pictorial content.” She provides the catalogue entry:

He is represented standing in his laboratory on the left side of the print. He wears a white cap and academical gown. His writing-desk lies before him on a table, on which his closed right hand is resting, while the other hand is placed upon the left arm of the chair, from which he has just risen, attracted by the sudden appearance of a luminous magic circle in the centre of a casement to the right. He is apparently watching the movement of a shadowy hand which points to a reflection of this circle in a mirror held by another hand below. Lower down, on the right, a pile of books lies on the table, and below, in the right corner, is the upper half of a globe. In the left background are a shelf of books, an hour glass, a skull, etc., while many sheets of paper fastened together hang by the upper part of the casement.34

Kraus goes on to compare Middleton’s description to a catalogue description of the same image made 100 years later as a supplement to a photographic reproduction of the print, and containing “no pre-iconographic information whatsoever.”

The pre-iconographic model of the catalogued image operates successfully within the current milieu of photographic critique, where the materiality and technicity of the image take on new importance and arguments around interpretation and signification are perceived as less urgent. The description locates and records the physicality of the visual content of the image, and does not set out to assign meaning to the objects with which it engages, because meaning that is based on intuition and uncorroborated by visual content is regarded as extraneous and even problematic in the archive. The description writer needs to present a neutral view that is open enough to allow different research directions, but which presents and supports the position of the archive image as information and evidence. Description is thus a text form that must be carefully managed. The archived photograph is a discrete unit, held in stasis on the shelves of the strongroom, its context dependent on surrounding objects, and according to its position and relationships recorded in the archive catalogue list. Images in the archive are therefore not considered autonomous objects that carry their own language, the much-peddled “universal language of photography” that is termed “bourgeois folklore” by Allan Sekula,35 a phrase which, notably, evokes a similar tone and sentiment as does Kenneth Goldsmith’s characterization of translation as “quaint, a boutique pursuit from a lost world.”36 Sekula argues vigorously that meaning is not intrinsic to the photographic image and explains: “it is clear that photographic meaning depends largely on context […] photographs, in themselves, are fragmentary and incomplete utterances. Meaning is always directed by layout, captions, text, and site and mode of presentation.”37 The prescribed approach of archival description demands that images take their cultural and spatiotemporal context from their “site and mode of presentation”; that is, from their place in the archive and the archive catalogue. The structure of the archive is the sign system for the archived photograph.


Performativity and Procedural Directives

Archival practices are performative in nature. Margaret Iversen sees performative practice as one that “begins with an instruction or rule which is followed through with a performance.”38 Her specification of two distinct actions, the second dependent on the first, clearly differentiate this use of “performative” from the early designation by J.L. Austin, in How to Do Things with Words, where the “saying” and the “doing” are one and the same thing (famously, the performative utterance of “I do” in the course of the marriage ceremony).39

Standards for archives were formally laid down over a century ago, in 1898, in The Manual for Arrangement and Description of Archives (commonly known as The Dutch Manual).40 Today, many archives take their guidance from the International Council on Archives’ ISAD(G): General International Standard Archival Description, a document that sets out clear rules and guidelines for writing and organising descriptions at every level.41 Though it is accepted that two different archivists’ descriptions of a given image would never match – personal experiences, approaches and habits materialise differences, for instance, and visual elements may be described in different orders and words – standards have been laid down to manage this variation.

Descriptions equalise: an element in the background may be as important to a researcher as one in the foreground, a small element as important as a large one, an object as important as a human. For instance, Kraus explains how, for Charles Henry Middleton, “there is no such thing as noise in the visual system that may safely be disregarded […] it is sins of omission that Middleton ultimately fears, rather than sins of commission.”42 The Dutch Manual emphasised the all-inclusive recordkeeping element of description: by describing and listing all elements, the cataloguer would leave the archive open to many uses and reconfigurations, rather than directing it towards any specific future use. This is a methodology that persists today.43 Accordingly, the description writer processes and presents details without making judgments on their importance in the past, present or future.

Judith Ellis quotes the words of eminent archivist Hilary Jenkinson, writing in 1948:

The Archivist’s career […] is one of service. […] His creed, the Sanctity of Evidence; his Task, the Conservation of every scrap of Evidence attaching to the Documents committed to his care; his Aim, to provide, without prejudice or afterthought, for all who wish to know, the Means of Knowledge.44

In latter years, there have been challenges to Jenkinson’s thinking by scholars of archival science who believe archivists should play a more interpretive role. Terry Cook, as part of his plan for the “postmodernisation” of archives, cites Foucault and Derrida, with their ideas on understanding and dismantling “systems of organised knowledge.” He calls for a change in archival practices to one that moves away from the guardianship of the document into a mode of “actively shaping collective (or social) memory” and for “a shift away from viewing records as static physical objects, and towards understanding them as dynamic virtual concepts.”45 Cook is, of course, writing in the context of expanding digitisation of archive material and the consequent fluidity of storage and call-up. Wolfgang Ernst’s views on the temporal dynamics of networked storage echo Cook’s, but Ernst cites his research year at the German Historical Institute in Rome as an insight into the workings of physical archives, “with their relational but not coherent topology of documents that wait to be reconfigured, again and again. The archival subject is thus a way out of the postmodern aesthetics of arbitrary anything goes – without having to return to authoritarian hermeneutics.”46 Even though the physical archive is preserved in stasis, with descriptions and lists systematically recording, presenting and protecting the fixed spatial relationships between objects, the physical archive emerges as a temporally dynamic space.

With regard to performativity and translation, Wilss suggests that systems of rules in translation could be developed to such an extent that creativity might likewise be done away with in that practice: “One could argue, e.g., that translation proceeds, at least for the skilled practitioner, in a routine fashion. This could mean that the concept of creativity, which is extremely controversial in the client/translator relationship, will go the way of phlogiston.”47 This is a radical position to take, but he goes on to explain how rules become internalised and are activated in the form of “rule-governed behaviour.” At any rate, it is clear that translation is carried out according to set rules and guidelines wherever possible.

Vinay and Darbelnet precisely outline seven procedures for translation that unfold with increasing levels of linguistic creativity and interpretation.48 As with the techniques for archivists set out in the Dutch Manual, their procedures were direct yet progressive when first formulated in 1958. Both these models have endured, against a background of new opinions emerging in their respective fields. Vinay and Darbelnet begin with the practice of “borrowing”: the “simplest of all translation methods,” a borrowing of the source language (giving the example “apparatchik”). Of procedure seven, “adaptation,” they write: “With this seventh method we reach the extreme limit of translation: it is used in those cases where the type of situation being referred to by the SL message is unknown in the TL culture. In such cases translators have to create a new situation that can be considered as being equivalent.” And they lament the absence of a set of rules for signification, “a conceptual dictionary with bi-lingual signifiers.”

While description and translation inhabit a space where creativity – in the guise of potential variance or divergence due to lack or neglect of rules – is controversial, they nevertheless continue to support the co-existence of personal approaches and adherence to rules and procedures. Like all activities that are hybrids of human and system, there is collaboration between technician and technique. As Geoffrey Winthrop-Young explains, “what we call the human is always already an emergent product arising from the processual interaction of domains that in time are all too neatly divided up into technical and human.”49 Hence Wilss’ notion of “rule-governed behaviour,” which forms the basis not only for communication systems such as description and translation, applies to wider societal behaviour systems as well.

Aside from the prescribed structures and non-human aspects that intervene in translation and in the description and classification of images, these still remain tasks best performed by humans, though there has been great progress in both machine translation and computer image recognition. Research into artificial neural networks has delivered exciting results in both fields, but the machine learning required for practical application (such as image description in the archive or translation of certain kinds of documents) is extensive and therefore not yet, or not always, cost efficient.


Dear Reader

In his critique of Georg Lukács’ “Narrate or Describe?” a 1936 essay that formalises the superiority of narration to description, Cannon Schmitt contests Lukács’ argument that “certain words in the pages of a novel are not actually pertinent to its meaning – to its narration or interpretation – because they are simply descriptive.” Schmitt quotes from Cormac McCarthy’s The Road:

And behind that was a composite toolbox, the opening of the lid sealed with black electrical tape. He pulled it free and found the end of the tape and peeled it off all the way around and unlatched the chrome snaps and opened the box. Inside was a yellow plastic flashlight, an electric strobebeacon powered by a drycell, a first-aid kit. A yellow plastic EPIRB. And a black plastic case about the size of a book. He lifted it out and unsnapped the latches and opened it. Inside was fitted an old 37 millimeter bronze flarepistol.50

Things dominate McCarthy’s passage. We can see through the Patrons description, too, that these things can be catalogued, recorded and, indeed, translated, in a direct and unambiguous way. For example, tables and windows are commonplace objects, placed within our knowledge of the everyday. They are also successfully spatially reframed with short phrases such as “on the table” and “towards the window,” which, again, lie within our acquired systems of knowledge, our “familiarity with objects and events,” in Panofsky’s words.51

According to Fowler, “[s]et-piece description is regularly seen by narratologists as the paradigm example of narrative pause, in the semi-technical sense of a passage at the level of narration to which nothing corresponds at the level of the story.” He explains: “The plot does not advance, but something is described.”52 And Wolfgang Ernst argues: “Description is at odds with narration.”53 Schmitt, though, suggests that one can interpret (or narrate) and describe, that the two are symbiotic, that descriptive elements in a text, in this case, “deployments of a technical maritime lexicon,” can be essential to the outcome of the plot and the lives of the characters, and that they serve “to lessen the distance between readers and characters.”54 Perhaps this is an appropriate moment to return to Barthes, whose first level of meaning is defined as one of detection, of knowledge-building and perhaps even of interpretation, as it “gathers together everything I can learn from the setting.”55

With respect to translation and the reader, Wilss argues that meaning is built through shared experience and does not need to be explicitly presented. He cites technical texts, already identified as suitable for direct translation, as an example, but goes on to apply this idea to a non-technical text, offering a descriptive passage from Alan Sillitoe’s novel The Loneliness of the Long-Distant Runner, followed by its German translation:

A middle-aged man wearing a dirty raincoat, who badly needed a shave and looked as though he hadn’t washed for a month, came out of a public lavatory with a cloth bag of tools folded beneath his arm.

Ein Mann mittleren Alters, der einen schmutzigen Regenmantel trug, eine Rasur dringend nötig hatte und aussah, als ob er sich einen Monat lang nicht gewaschen hätte, verlieβ, eine Stofftasche mit Werkzeug unter dem Arm tragend, ein öffentliche Toilette.56

The German translation is direct. It is open to the reader to apply their knowledge of the world in order to identify the situation described. One might say that the description of the man is evidence that he is a worker, perhaps poor and homeless. This is an intentional linguistic trope by Sillitoe, and the translator has understood and respected the author’s intent, thus providing a direct and rule-based translation of information that could be relevant to the plot. However, some creativity of thought is involved in the translation: as Wilss observes, “the translator imagined the reported event just as Sillitoe had.”57

Similar situations materialize in the Patrons text, for example, in the description of the man in the corner of the cafeteria:

At the furthest table, on the chair nearest the back wall, sits a man with dark skin, a long grey beard and a receding hairline. He is wearing a dark grey casual coat, light coloured trousers and black and white trainers. There is a large blue bag of some kind placed on the chair opposite him. Both his hands are placed on the table and he appears to be eating a meal from a large, red tray. The rest of the table is covered with what could be packaging or paper and also a drinks can.

And then later:

Both figures are looking towards the action that is taking place on the other side of the window from them. The man in the corner is looking in a guarded and surreptitious way, his head kept straight and only his eyes slightly turned. The man at the middle table is looking with unconcealed interest, his head and eyes lifted from the newspaper and turned directly towards the window.

One might conclude that the man in the corner is also homeless, that he is taking a meal in this evidently down market eatery (“The floor is covered in small beige tiles that extend slightly up the wall in a functional style”) and that all his worldly possessions are in the bag on the chair opposite. He is looking furtively at the events taking place outside, perhaps because he does not want to draw attention to himself and his situation, unlike the apparently more well-to do man sitting at the middle table, “neatly dressed in a pale pink sweater and beige trousers,” who is looking with unconcealed interest.”

Benjamin argues against “consideration of the receiver” in translation and beyond, and finds the concept of the ideal reader damaging, since it makes assumptions about the “nature of man.”58 Whilst the Patrons text is written within Panofsky’s second level of meaning (“a familiarity with specific themes and concepts”), and without an ideal reader in mind, accumulated and shared knowledge amongst different readers might cause it to stray into the third level, a “familiarity with the essential tendencies of the human mind.” The deductions I have made, outside of the description, about the status of the men in the restaurant (and the man in the Sillitoe extract), are made as reader and although other readers may similarly extrapolate, there is a possibility that an entirely different storyline exists, one that does not fit so well with collectively understood human behavior patterns. That is why I cannot make these or any other deductions as writer. Victor Burgin begins his book The Remembered Film with a Wittgenstein quote, which concludes, “If you complete it you falsify it.”59 This is valuable advice for translators of all kinds.



Wilss considers translation “essentially a “derived” linguistic activity.”60 “Derived,” in Wills’ characterization, means that “the purpose of translation is not the creation of an authentic text, but the transformation of a primary text into a secondary text.” Image description is also derivative in this sense as it, too, originates from a source image. Authenticity and creativity, it would seem, are thus irrelevant. The use of the words “primary” and “secondary,” which are standard expressions in translation for source and target text, might be perceived as denoting value, suggesting that the source text is more worthy than the target. When it comes to image and description, however, these terms are purely sequential and I maintain equality between the two. For translation, Benjamin sees this sequentiality, the necessary time lapse between the production of source and target texts, as critical to the “afterlife” of the text, arguing that “translation marks their stage of continued life.”61 The archive respects a similarly conditional temporality. There, objects may be described soon after their production, or many decades later. Such descriptions are frozen in time by the cataloguing process but, at the same time, are given “afterlife” by always existing in the present, what Peter Wollen terms “the spectator’s “now”.”62

Text description is perceived by D.P. Fowler as the superior medium: he remarks of Sciascia’s 1912+1 description, “there is a obvious sense in which description in language inscribes a point of view more forcefully and more unambiguously than plastic art.”63 Fowler is right that a certain kind of clarity can materialise in the language of description that is sometimes lacking in the image, mainly due to the equality of description of visual elements that materialises through the record-keeping process. Take, for example, this description of a photograph from the Mountbatten Archives, in the University of Southampton Library’s Special Collections:

Black and white photograph of a busy street scene in Luxor. Only the buildings on one side of the street are pictured. They are simple buildings, one or two storeys high, made of mud bricks and flat roofed. Many of the windows are without shutters. In the foreground there is a single story extension on the front of a building, with a grass thatched roof. Outside this extension there are many earthenware pots. Most notable are large flagons with pointed bases, which are leant against the side of the building. Many of these flagons are propped on the roof of the neighbouring house. There is an extension at the front of this building as well, with an awning extending into the street. This is also being used as a shop, but its wares are concealed by a group of men, including two men with laden donkeys. There are many more men in the street, all wearing kaftans and turbans. In the distance a minaret shows above the other buildings.64

It is also important to consider the controlled temporality of reading descriptions such as this: one is locked down during the process of reading a detailed description in a way that does not necessarily happen in the context of viewing an image. Image viewing is often a brief and desultory act as the viewer quickly scans for something they might expect to find, but not for what is actually there. Again, Marcus et al insist that recent, positive evaluation of description urges us “to consider it on its own terms and not as a stepping-stone on the way to interpretation and critique.”65

Robbe-Grillet explains that his “cine-novel,” The Immortal One (a semi-technical translation from film to text, or, more accurately, from script to imagined film to text, as the text preceded the film), can be read without seeing the film, “in the same way as a musical score; what is then communicated is a wholly mental experience, whereas the work itself is intended to be a primarily sensual experience, and this aspect of it can never really be replaced.”66 Working with image description is working on the borders of the experiential, between sensational image and seemingly dry text. The act of describing becomes a vehicle for the examination of the photographic image that is beyond traditional representation and hermeneutics, but the description itself, in spite of its being a straightforward, evidence-based form, can be poetic, can embody certain aesthetic qualities, although these qualities are quite different to those embodied in the image itself.

While Benjamin rejects the notion that great poets automatically make great translators, asserting that they practice quite distinct tasks, he still places translation “midway between poetry and doctrine.”67 Description and translation continue to share this curious space.


  1. Wolfram Wilss, Knowledge and Skills in Translator Behaviour (Amsterdam/Philadelphia, PA: John Benjamins, 1996), 105. 

  2. Jean-Paul Vinay and Jean Darbelnet, “A Methodology for Translation,” in The Translation Studies Reader, ed. Lawrence Venuti (New York: Routledge, 2000), 86. 

  3. Wilss, Knowledge and Skills in Translator Behaviour, 59. 

  4. HEADLINE: Patrons watch an activist banging on the window of McDonald’s in Los Angeles. CREDIT: Mario Anzuoni. SOURCE: Reuters. 

  5. Bruno Latour, Reassembling the Social: An Introduction to Actor-Network Theory (Oxford: Oxford University Press, 2005), 127. 

  6. Latour, Reassembling the Social, 137. 

  7. Lawrence Venuti, Translation Changes Everything: Theory and Practice (New York: Routledge, 2013), 110. 

  8. Venuti, Translation Changes Everything, 110. 

  9. International Council on Archives, ISAD(G): General International Standard Archival Description (2000), 12. 

  10. International Council on Archives, ISAD(G) ), 12. 

  11. D.P. Fowler, “Narrate and Describe: the Problem of Ekphrasis,” Journal of Roman Studies, 81 (1991): 29. 

  12. Kari Kraus, “Picture Criticism: Textual Studies and the Image,” in The Cambridge Companion to Textual Scholarship, ed. N. Fraistat and J. Flanders (Cambridge: Cambridge University Press, 2013), 242.  

  13. Fowler, “Narrate and Describe,” 26-27. 

  14. Hito Steyerl, The Wretched of the Screen (Berlin: Sternberg Press, 2012), 32. 

  15. Walter Benjamin, “The Task of the Translator: An Introduction to the Translation of Beaudelaire’s Tableaux Parisiens,” in The Translation Studies Reader, ed. L. Venuti (New York: Routledge, 2000), 15. 

  16. Kraus, “Picture Criticism”, 239-241. 

  17. Wolfgang Ernst “Archive in Transition,” in Interarchive, eds. B. Von Bismark, H.-P. Feldman, H.U. Obrist, D. Stoller & U. Wuggenig (Köln: Verlag der Buchhandlung, 2002), 482. 

  18. International Council on Archives ISAD(G), 11. 

  19. More about the Library of Congress Flickr project here:

  20. Stephen Connor (2002) “Sleights of Voice: Ventriloquism, Magic and the Harry Price Collection” [Online]. Available: [accessed 12-11-12]. 

  21. Vinay and Darbelnet, “A Methodology for Translation,” 86-87. 

  22. Sharon Marcus, Heather Love and Stephen Best, “Building a Better Description,” Representations 135/1 (2016): 6. 

  23. Fowler, “Narrate and Describe,” 27. 

  24. In A. Halsall (1988) ““La Transition”, Description et Ambiguités Narritivo-Discursives dans “Victoire” de William Faulkner,” in L’Ordre du Descriptif (n.2), ed J. Bessière (Paris: PU de France), 159-72. [original italics] 

  25. Georges Perec (2011) Things: A Story of the Sixties (London: Vintage), 24. First published in French under the title Les Choses (1965). 

  26. Benjamin, “The Task of the Translator,” 18-19. 

  27. Wilss, Knowledge and Skills in Translator Behaviour, 52-53. 

  28. Wilss, Knowledge and Skills in Translator Behaviour, 54. 

  29. Wilss, Knowledge and Skills in Translator Behaviour, 55. 

  30. Federica Scarpa, “Closer and Closer Apart? Specialized Translation in a Cognitive Perspective,” in Translation Studies: Perspectives on an Emerging Discipline, ed. A. Riccardi (Cambridge: Cambridge University Press, 2002), 142. 

  31. From Kenneth Goldsmith’s Facebook page: [accessed July 2016]. 

  32. Erwin Panofsky, Studies in Iconology: Humanist Themes in the Art of the Renaissance (New York: Harper Torchbooks, 1962), 3-17. 

  33. Roland Barthes, Image, Music, Text (London: Fontana Press, 1977), 52-55. 

  34. Kraus, “Picture Criticism,” 241-242. 

  35. Allan Sekula, “On the Invention of Photographic Meaning,” in Thinking Photography ed. V. Burgin (Basingstoke: Macmillan, 1982), 86. 

  36. Kenneth Goldsmith, “Displacement is the New Translation’, Rhizome (9 June 2014) [accessed 21 June 2016]. 

  37. Allan Sekula, “Reading an Archive: Photography Between Labour and Capital,” in The Photography Reader, ed. L. Wells (London: Routledge, 2003), 445-446. 

  38. Margaret Iversen, “Auto-Maticity: Ruscha and Performance in Photography,” in Photography after Conceptual Art, eds. D. Costella and M.E. Iversen (Chichester: Wiley-Blackwell, 2010), 15. 

  39. J.L. Austin, How to Do Things with Words (Oxford: Clarendon Press, 1962), 5. 

  40. Samuel Muller, R Fruin and Johan Adriaan Feith, Manual for the Arrangement and Description of Archives (translation of the second edition) (Chicago: Society of American Archivists, 2003). 

  41. International Council on Archives, ISAD(G)

  42. Kraus, “Picture Criticism”, 245. 

  43. Eric Ketelaar, “Archival Theory and the Dutch Manual,” Archivaria 41 (Spring 1996): 33. 

  44. Judith Ellis, Keeping Archives (Port Melbourne, Australia: D.W. Thorpe, 1993), 12. 

  45. Terry Cook, “Archival Science and Postmodernism: New Formulations for Old Concepts,” Archival Science 1 (2001): 3-6. 

  46. Wolfgang Ernst, Digital Memory and the Archive (Minneapolis: University of Minnesota, 2013), 194. 

  47. Wilss, Knowledge and Skills in Translator Behaviour, 145-146. 

  48. Vinay and Darbelnet, “A Methodology for Translation,” 84-93. 

  49. Geoffrey Winthrop-Young, “Cultural Techniques: Preliminary Remarks,” Theory, Culture and Society (Special Issue: Cultural Techniques) 30, no.6 (2013): 10. 

  50. Cannon Schmitt, “Interpret or Describe,” Representations 135/1 (August 2016): 109-112. 

  51. Panofsky, Studies in Iconology, 5. 

  52. Fowler, “Narrate and Describe,” 25-26. 

  53. Ernst, Digital Memory and the Archive, 153 [original italics]. 

  54. Schmitt, “Interpret or Describe,” 106-110. 

  55. Barthes, Image, Music, Text, 52-55. 

  56. Wilss, Knowledge and Skills in Translator Behaviour, 161. 

  57. Wilss, Knowledge and Skills in Translator Behaviour, 161. 

  58. Benjamin, “The Task of the Translator,” 14. 

  59. Victor Burgin, The Remembered Film (London: Reaktion Books, 2004), 7. 

  60. Wilss, Knowledge and Skills in Translator Behaviour, 175. 

  61. Benjamin, “The Task of the Translator,” 16. 

  62. Peter Wollen, “Fire and Ice,” in The Photography Reader, ed. L. Wells (London: Routledge, 2003), 76. 

  63. Fowler, “Narrate and Describe,” 29. 

  64. MB2/L4/6 Black and white photograph of a busy street scene in Luxor, c.19 January 1928 – 10 February 1928. University of Southampton, Special Collections. 

  65. Sharon Marcus et al, “Building a Better Description,” 3. 

  66. Alain Robbe-Grillet, The Immortal One (London: Calder & Boyars, 1971), 5-6. 

  67. Benjamin, “The Task of the Translator,” 20-21. 

Article: Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported.