The Photograph in Theory

by Elizabeth Chaplin
Open University

Sociological Research Online, Volume 10, Issue 1,

Received: 28 Jul 2004     Accepted: 10 Nov 2004    Published: 31 Mar 2005


This article gives an account of an ethnographic project which relied on the use of photography: the project involved taking photographs, asking for responses to them, and then analysing both photos and responses. Anything that plays a central role in an ethnographic project requires theoretical consideration. So, as the account of the project proceeds, the photos are considered from different theoretical viewpoints; and an emerging and subsequently recurring theme is the tension between what a photo shows and what a photo means. Discussion of this tension develops into a more general critique of the ways photos are theorised in social science. I conclude that the photograph in social science theory is at present a sad phenomenon, and that in order to remedy this situation, we should seek help from outside social science.


1.1 This paper starts by reporting an ethnographic project based on the use of photography: the project involved taking photographs, asking for responses to them, and then analysing both photos and responses. Anything that plays a central role in an ethnographic project requires theoretical consideration. So, as the account of the project proceeds, the photos are considered from different theoretical viewpoints; and an emerging and recurring theme is the tension between what a photo shows and what a photo means. Discussion of this tension develops into a more general critique of the ways photos are theorised in social science. I conclude that the photograph in social science theory is at present a sad phenomenon, and that in order to remedy this situation, we should seek help from outside social science.

1.2 In May 2000 I set out to make a systematic photographic record of all the residents in the road where I live.

In my introductory letter to each of the 36 households, I said I wanted to photograph everyone standing outside their front door, and that my aim was to capture 'a slice of life in the year 2000'. Just two households declined to take part in the project; a further four allowed me to photograph only their front doors

and all the rest were happy to co-operate fully

The six complete or part-refusals included the three families in the road of Asian or part-Asian origin, but also three White families who for various reasons did not want visual evidence of themselves or their house-front to enter the public domain. [1] I kept a field diary for the duration of the project, and for a while was updating this on a daily basis.

Mine wasn't the only millennium photographic project involving UK residents. But it differed from the others in being an academic study, a piece of ethnographic research carried out for the National Everyday Culture Programme at the Open University. [2]

1.3 Photos have long been used in ethnographic research because they have an immediacy which words can rarely match; and because the sheer visual detail in the 'frozen moment' repays repeated scrutiny, since it may yield clues that the researcher finds profitable to follow up.[3] For example, at No.18, close inspection of the photo reveals that those two shirts are identically patterned, one brown, the other blue,

thus hinting, perhaps, at a rather closed type of family relationship - and possibly opening up an avenue of research: would an examination of all the South London Road photos yield clues about different 'family types' in existence at the turn of the millennium? Following up this line of research could affect the focus of the project, and in one respect it did.[4] But my concern in this paper (and it increasingly became my intellectual concern in 2000) is how to account theoretically for the photos within the research project. To start with, we can say that if a photo provides evidence of something, we are talking about an empirical approach based on an assumption of some kind of physical relationship between reality and a photo of that reality. I shall call this approach 'photographic realism'. But stepping back to another - cultural - level, we are dealing with a belief: a belief in the photo as evidence-of-something. This belief is still widespread, to judge by the comments of the residents,[5] and is therefore a powerful cultural force in its own right; akin for example, to the common belief that motorway surveillance cameras can record the speeds of individual motorists and thus justify the imposition of hefty fines on them. At this cultural level - the level of the discursive - the project seems to demand a Foucauldian approach. The next step might be to ask which is the more significant phenomenon, the demonstration of objective evidence, or the force of the belief? Which level should the project be focussed on? Or could the two theoretical approaches be brought together? I attempted to make a theoretical bridge between the two positions by comparing closely the work of Erving Goffman (1979) and Stuart Hall (1973). The results of this attempt are not without interest (they influence what follows on p.7); but eventually I decided to leave the theoretical tension in place, and to explore each level in turn.

1.4 When the photography was done and the negatives processed, I delivered a package to each set of participating residents. This package contained a copy of their photo, as originally promised, and also a letter thanking them for co-operating and asking a further favour: would they write a caption for their photo, and deliver it through my door? A third of the households did so: a few captions were supplied by men on the spot, but most were posted through my door by women later, probably out of a sense of duty. These captions turned out to be diverse in character, and made a considerable difference to one's previous grasp of the image-content,

Figure 3. Motherless Boys

Figure 5. The Survivors

Consequently, I began to shift attention from a) what the photos show, and b) the force of the belief that photos do not lie, to c) the influence of the caption (plus talk and general contextual baggage) on the viewer's interpretation of the photo. So to recap: I had started the project focussing on the photo as record (photographic realism); I then saw that the idea of 'the photo as record' could itself be the object of study; and now I was treating the photographic image as polysemic - the site of shifting meaning. Meanwhile, the residents themselves were having no apparent difficulty in treating their photo as both visual record and as open to interpretation, thus demonstrating the cultural force (and apparent compatibility) of each of these beliefs. In my subsequent theoretical analysis of the research material, therefore, I treated the photos first as visual records, secondly as texts-to-be-read, and thirdly as the focus of a widely held belief. The second and third both entail a broadly constructionist approach; with semiotics (the second) focussing on how the meaning of a text is produced and conveyed, and Foucauldian theory (the third) being a more generalised conception of constructionism concerned with how widely held beliefs - including truths and norms - come into being, and are then incorporated into cultural technologies for the management of social order.

1.5 Let us start by analysing the photos as visual records of the residents. What theoretical precedents are there for treating photographs empirically as records, and how convincing is the theory underlying this approach, that of photographic realism? Within the empirical approach there seem to be two alternative avenues: either analyse the photographic image in its own right, following Goffman (1979); or treat the image-plus-caption as an information package. This latter 'method' (I use inverted commas because its theoretical basis is only rarely examined by its proponents) has been commonly used, for example, in museums where the caption is treated as confirming information in the image and supplying additional facts. Now clearly, some of the residents' captions did contain a lot of extra information. For example the caption for No 6 says:

The photograph shows Mrs Nellie Horton (born 31/12/1914) and her son, Dr Nigel Horton. The family purchased No.6 in 1968, and by 2000 were among the longest standing residents of South London Road. The front of the house, including the front door, is basically as built, although the railings and gate (not visible) are 1970s replacements for those removed in the scrap metal drive in the Second World War.

Dear Elizabeth, (20 South London) Occupants - Vicky and Paul (New Zealander) , children: Sam (7) Harry (15 months) Hoping to move to New Zealand in the future to enjoy a more simpler way of life, to get away from the stresses of London.
We have lived here 8 years and although we enjoy living here we feel we would like to move on. We would like our children to grow up nearer the sea and countryside.
(If possible may we have 2 further copies of the photograph you kindly took. The one with all 3 of us on) Thanks.

1.7 So some captions do clearly supply additional information about all sorts of things - both inside and outside the picture, as well as confirming items in the picture itself. My set of images-plus-captions would probably find a place in the local authority's history archive. But even the most factual-sounding captions are not just add-ons to the image. They are never neutral; even what they leave out is significant, because the omission indicates that something in the picture was not considered important enough to mention. Whether intended to or not, captions transform images, give them particular meaning, mould our interpretation of them - and thus provide grist to the constructionist mill.

1.8 Perhaps it was the knowledge that captions are an unwelcome complication for the empirical purist that caused Erving Goffman - in pursuit of theoretical clarity - to attempt a different and apparently unique method of analysing photographs. In Gender Advertisements (1979) he brackets any captions or writing on the photographic image, and focusses his sociological attention on the image itself, This seems like a test case concerning 'the autonomy of the image'. Can pictures be made to 'speak for themselves'? Can they remain detached from the interpretative grasp, the suffocating tentacles of words? The last chapter of Gender Advertisements contains 500 photos divided into categories; each category being devoted to a different type of display of gender inequality. I tried fitting my photos into these categories, for example:

  1. Border control: The man adopts a non-sexual protective stance, with his arm round the woman; the inferior may snuggle into the superior:

  2. Or, in the case of a woman and children: The woman has a protective arm on or round the child:

    The protective arm disappears as the children get older - remember motherless boys?)

  3. Socially inferior members sometimes act cutely, and do clowning:

  4. The woman often smiles; the man doesn't. The smile is the offering of an inferior: she wants to be a member of the group at hand; but he does not have to make the effort:

  5. In a collaboration between a man and a woman, given that there is one executive role, the man takes it:

1.9 Now most of Goffman's images show a man and a woman, a family group, or one rather glamorous young woman; and they are mainly advertisements. But mine are the products of a different context, and indeed a different date - thirty years later; and some of them seem to call for an extension or modification of Goffman's categories:

No 1 South London Road could be seen as an extension: While the man in front (the resident) takes the executive role, the one behind (his friend) refuses the role of an inferior. And

at No 9, she is inside the door for protection, arguably because there is no man present. But then

this single male also stands back inside the door frame - though admittedly his folded arms and smile suggest confidence. And are these

examples of the woman smiling while the man does not? And so on. It can be difficult applying Goffman's categories; but is this because his images are a judiciously selected sample of 'gendered' advertisements, while my photos constitute almost an entire population? Apparently not entirely, for, as Greg Smith has pointed out (1996), category exceptions and dubious examples abound even amongst Goffman's own selection of images.

1.10 In addition, Goffman's overall theoretical approach to photos is not what it seems on the surface. In one of the dense verbal essays which precedes the final chapter containing his pictorial pattern analysis, Goffman argues that the reason why we can talk about what a photo shows is because people from the same culture interpret a given image similarly: our previous shared cultural experiences incline us to give it the same basic meaning. In other words, the photo-as-visual-record is underpinned by social constructionism. And indeed, the way his text and images are laid out in the final chapter of Gender Advertisements is a tacit acknowledgment that the meaning of an image can never be pinpointed with absolute certainty. Each page is organized in vertical columns. At the top of the left hand column is a category heading - a named principle of gender display, for example, 'licenced withdrawal'. Then come one or two very dense paragraphs telling us what to look out for, and then finally come the images. Thus Goffman makes sure we refine our basic shared interpretations of the images along his preferred lines before actually viewing the images themselves. In other words, he turns out to be a closet constructionist, after all.

1.11 In summary, both attempts to analyse the photos empirically turn out to be theoretically shaky. The caption may add 'information' to that contained in the photo, but it is never neutral information, which makes the term 'information', and even 'data' seem rather unsatisfactory. What first appeared to be straight information now seems bound up with values and interpretation. Precisely the same comment can be made about Goffman's work although, as a theoretician himself, he does very briefly acknowledge that his approach is underpinned by the cultural phenomenon of consent. Let us now turn, therefore, to constructionism.

1.12 My second analytical approach is overtly constructionist, and treats the image and caption together as a text-to-be-read. It therefore contrasts with photographic realism, where the caption is either treated as supplementing the information in the photos, or is virtually ignored in favour of a sociological analysis of the image in its own right. Focussing now on a semiotic approach within the overall constructionist paradigm, it is the relationship between the image and text which concerns us. The image is regarded as polysemic - i.e. its meaning would float if it were not for the caption, which does the job of establishing a specific meaning for it. Stuart Hall (1973) has suggested that what actually happens is that the caption selects one from a range of possible potential meanings for the image, and then amplifies it , and it is this formula that I will now apply to my own data.

1.13 In Fig. 9

Figure 9. The House before Marriage

if we cover up the caption, we see a woman and a man close together, in front of a house. But, uncovering the caption - 'The House before Marriage' - it tells us that these two live in the house together, but are not yet married. And now the image comes into its own again. With the help of the caption, we now view these people according to the wider ideological theme of 'modern living'. We see two confident-looking people who are rejecting yesterday's values in which marriage came before buying a house. Everything about these two now seems to suggest they are promoting today's social norms and values. This reading of the image seems like factual information obtained from the photo itself. But what is happening is that the ideological work of the caption is being absorbed into the photo and disguised as visual denotation.

Roger and Carole Owen are our names
Train driver and Legal Receptionist are our games

We've been here since 75
Children none - cats five!

Shops were plenty - now are few
Cars have increased considerably too

District and Main Line trains are to hand
And all our neighbours have generally been grand

So we have stayed here all this time
Simply because it has suited us fine

(and we have liked it here)
Carole Owen.

1.14 When viewed in conjunction with this remarkable offering, the image signals that either Mrs. Owen is a rather unusual legal receptionist - or that legal receptionists are very imaginative people. But Mrs. Owen stressed to me that 'we are very ordinary, really', which shows that the image's meaning isn't shaped by the caption alone, but also by the surrounding verbal context in which it is viewed. In marked contrast to the captions provided by Mrs. Owen and and the residents of no. 4, is the caption for no. 6 (see Fig. 6) which I shall now re-examine. This caption presents the photo as a serious, systematic, historical document. It describes in a plain manner items that can - and cannot - be seen in the photo, and backs this up with further historical facts. Looking back now at the photo, it seems apparent who wrote the caption. It is the man, Dr. Nigel Horton, who does now seem to have the air of a scientist-intellectual, which gives his stance - his protection of the older woman - an appropriate gravitas. Again the ideological, connotative work of the caption - in this case to do with science, men and serious information - is absorbed into the photo as fact; although Dr. Horton probably intends his caption to be an addition to the image, the photo-as record approach, rather than pinning down one reading of it.

1.15 In summary, this semiotic approach is useful in explaining how the caption affects our interpretation of the image. Hall's method, when applied to my data, yields what seems a fruitful analysis. Yet a criticism of semiotics is that its findings cannot be verified. The above analysis might not be replicated by another researcher. It is mine, and someone else might have picked out different aspects of the image and caption to link up and discuss. However, that inbuilt weakness of semiotic theory is in a sense minimised when the micro-constructionist approach (i.e. semiotics) is theoretically linked to and supported by a macro-constructionist enterprise (Foucauldian theory). Which brings us to the third way of analysing photos and captions, namely that of theorising them as part of a widely held belief. Like the other residents, Dr. Horton took for granted that 'his' photo was an accurate record of what it depicted. After all, a photo can count as evidence in a court of law. But the

theory supporting the photo-as-record is unconvincing, as we have seen. Even Goffman (1979) mentions briefly that we are able to talk about what a photo denotes because at a basic level we are in agreement (that it contains e.g. a house, a tree, two people, a dog..). So perhaps all that is happening is that people believe that photos are reliable records. Here we are close to a Foucauldian approach. Foucault argued - contra Marx - that what is believed, what counts as the truth, as knowledge, even as reality, is relative to the culture which generated those beliefs and truths. Each culture has discursive formations, the key participants in which, usually professionals occupying key posts in social institutions, create - through thought, talk and action - discourse: that which percolates through to the general population, and is held to be correct and true. Foucauldians might, for example, argue that, in Western culture, photographic realism was helped to prominence by the sheer dominance of the discursive formation of positivism when photography was in its infancy in the 19th century; and that photographic realism consequently proved effective as an instrument of social management, for example, in courts of law. They might add, though, that the 21st century intellectual's disdain for positivism has been slowly fuelling a growing suspicion of 'the facts' via television history programmes and school curricula [6] which, in conjunction with the rise of digital cameras, may eventually get the ordinary person abandoning the idea of the photo-as-record.

1.16 Seen in retrospect, my project shows signs that this might be happening. Though the South London Road photos were taken with an analogue camera and I told the residents I was making a photographic record - and most of them followed me in treating the photos straightforwardly as evidence - there was an exception. At least one resident knew he could play tricks with the caption, using it to gently tease viewers into supposing that those boys had no mother (they usually had a full-time one, but she was, exceptionally, away in Australia attending her father's funeral). So one resident was chipping away at the dominant discourse of photographic realism. And in fact by the date of the research (2000), the prints themselves could quite easily have been scanned and the images altered digitally. (As a late-middle-aged amateur photographer - i.e. as definitely not a leader in the field - I was already scanning my images into the computer.) Indeed, we were being informed that newspaper photos were not necessarily any longer reliable: there had been several articles in the press, complete with 'before' and 'after' image, pointing out that if the photo in question had remained undetected, this would have given a false impression of current events and boosted both the reputation of the photographer and the sales of the newspaper which published it.[7] In sum, the concept of documentary photography was beginning to look shaky in 2000. And in the period between then and now (2004) the sale of digital cameras has risen so that they far outstrip sales of analogue ones,[8] while advertisements for software called 'editing suites' have begun to appear in the national press, thereby increasing the general suspicion, and indeed, likelihood, that photographic images are no longer what they seem. If I embarked on the South London Road project now, I would be using a digital camera. But would I still claim to be making a photographic record? Foucault's account of the birth of modern medicine in the early 19th century describes how one discursive formation took over from another; a situation in which 'the deciphering of hidden essences' was replaced by 'the description of symptoms' (Foucault 1973a, p.137). From a Foucauldian point of view, are we looking at a historically specific instance (the South London Road project, 2000) of a discursive formation in the making? While the new digital photography (ideas, camera technology, talk, action) was not widely established in 2000, the present incumbent of the dominant photographic discursive formation - photographic realism - no longer seemed totally secure. A new discursive formation looked set to take over.

1.17 But will it? Stepping back, as it were, from Foucauldian theory, we can see that it draws a line round photographic realism, as it does round other 'discursive formations'. It relativises them, in part because the theory itself actually proposes a relativist view of the world (as opposed to an essentialist one), but in part by the mere fact of claiming the main theoretical focus for itself. But photographic realism resists being forced out of the main focus. It is not easily relegated to second place. The Daily Mirror in May 2004 contained photos which were said to show British soldiers torturing Iraqi citizens. These images looked unconvincing to most people and were fairly quickly established as fakes; which indicates that the digital photographer of average competence may be able to alter images convincingly (eg. airbrush someone out of a group and replace them with indeterminate cloud and sky - see Fig. 25); but that building a realistic-looking photographic image from scratch of something highly specific and involving complicated actions, such that it will convince viewers that what is portrayed actually took place, remains extremely difficult, if not impossible. This suggests that photographic realism will be with us for some time, and remains a force to be reckoned with.

Figure 25. 'From the Guardian Weekend 5.6.2004'

1.18 And there are other situations where photographic realism is difficult to dismiss. In my project, one of the households refused to participate because they feared that if they were photographed, someone they were trying to avoid might see the image, find out where they lived, and come and do them harm. At your peril do you ignore the fact that photos can provide evidence. They can. Even cultural studies scholars routinely regard photos as evidence when they are not in theorising mode (see my Fig 24!). When they write articles and books which include images or diagrams, they are presumably assuming that these images and diagrams are pretty accurate photographic reproductions of 'originals', and also that the article or book in question will be reproduced accurately each time it is printed. And the reason for making this assumption - as it is in all cases where visual evidence is an issue - is, of course, that taking a photo with a camera involves automatic mechanical, electronic and chemical processes which are outside the photographer's control. If a photo is taken by accident, without the intention of the photographer, a picture results. Even if the result is intended, the photographer can learn, by looking at it, of things that s/he did not know were there at the time the picture was taken. When the light passes through the lens it leaves a trace on the treated sensitive film. That is a non-humanly produced trace of the world out there.[9]

1.19 If photographic realism exerts such a force, why has theorising it proved such an uphill, dispiriting task? Do we really have to accept that 'photographic evidence' is a matter of cultural agreement - something that can be explained in detail as a dialectical relationship between image and caption? The dominance of constructionist theory (in Britain at least) has obliged those who would put the case for photographic realism (e.g. Hamilton, 1999, pp.125-6) to tread very carefully indeed around the established credentials of constructionism. In addition, there is a political factor. Via its association with passť positivism, photographic realism is associated with conservatism - in contrast to constructionism with its connotations of the (semi-) autonomous and even creative individual. And this association is one that social scientists sympathetic to photographic realism may be reluctant to take on, given the general critical character of social science itself.

1.20 But the basic problem is that theory is not congenial to photographic realism (to the autonomous image). Constructionism brackets the image as record, and emphasises that images are polysemic, and this means that words are needed to choose and tie down the meaning of the image. And once the image is, as it were, defined by its caption, the door is opened for theory to enter, and for the image to be surrounded by words and treated as part of a verbal world. But, by contrast, photographic realism brackets the polysemy of the image. Instead it emphasises that the image is a record, the rationale for this being that what defines the camera as a camera is the fact that its mechanism allows it to produce a trace (either on light-sensitive paper or electronically as pixels) of the reality it has recorded. This trace is at the heart of the argument for photographic realism - and because the focus is on its physical presence, a semantic no-go area opens up between it and any verbal theory that tries to get near it. There's no suggestion of polysemy here: a trace is a trace (unless either you can prove it isn't, or you can convince people that 'trace' might not be the most accurate word to describe it...).

1.21 This theoretical impasse with regard to the photographic trace helps explain why in the world of academe (as opposed to the world 'outside' [10]) visual representation leads a rather impoverished existence. For in universities what counts as prestigious work - producing sound argument and theorising - is achieved by verbal means. Theoreticians are skilled wordsmiths. Photographic images are rarely used in theorising [11] (indeed this idea only seems to make sense with difficulty); rather they are themselves treated as part of the data to be theorised. Little wonder, then, that theoreticians are drawn towards a constructionist approach. Focussing on the polysemy of the image (the very idea of polysemy seemingly rooted in the verbal) allows theory to gain a foothold; and once this happens, the image looses semantic autonomy; that is to say, it is relegated to the role of illustration. Whereas to focus on the photographic trace is to come up against a situation where what words can seemingly most helpfully do is engage in halting description (that falls well short of theorising) - not a highly rated academic pursuit. So, we can see that the photograph exhibits a tension which is at the heart of social science: the polysemic image (with its sense of incompleteness without the support and clarification of words) as against the visual record, the physical trace, of reality which keeps its autonomy by forcing words into the subordinate role of description. Meaning as against showing.

1.22 Put rather differently and in rough terms (though these are terms which Peircian theory can help to flesh out[12] ) we could say that the more the focus is on the photographic image, the less room there is for focus on the (verbal) theory, and the more incoherent is that theory. Conversely, the more the focus is on theory, and on theoretical coherence, the more the image slips out of sight. Admittedly, and now making light of the tension between meaning and showing, it is possible to take a practical stance, and to argue that the photographic trace is a record of something which could have been recorded differently; and is, furthermore, open to many different interpretations. Indeed, one could argue that the degree of objectivity and subjectivity involved in each photo varies from image to image. Alternatively, as we saw earlier on, it is possible to step back from the impasse and to ask what part the photo plays in a particular discourse (for example, the discourse of motoring and road safety). This Foucauldian approach does at least require the photo to be scrutinised, but the visual content is not what is directly theorised, as the discussion of the ethnographic project showed. Words are better at the twists and turns of argument than images are, and once words themselves become part of what is to be theorised, they gradually take over and the image recedes into the background. By the time a second or third response to an article like this is made - a second or third layer of verbalising - the image has receded into being an illustration of a point being made, and ultimately exits entirely, leaving only words responding to words. And it is in order to highlight this tendency for the image to be caught in a downward spiral and in the end snuffed out that I have found myself fighting the corner of photographic realism. As someone who has been surrounded by images for the last 40 years, I value the semantic autonomy which photographic realism grants the photographic image, despite - what I consider to be - the lack of adequate verbal theory to back it up.[13] I believe it is important for social science to hang on to the photographic image, and to give it more autonomy than constructionism allows.

1.23 Taken to its 'logical conclusion', 'the answer' to 'the problem' (of theorising the photograph) might seem to be to produce visual theory along the lines of Constructivist art practice [14] , or even some cutting-edge science. Many natural scientists seem to explain new situations in visual terms, but these images and diagrams tend to be replaced by numbers and letters as the theory becomes more established and settled. [15] Constructivism claims to be scientific and to produce visual theory in the form of artworks [16] , and this can seem perfectly plausible within a fine art context; but outside that context, most people have difficulty in conceiving of theory as visual at all. And this is the final problem. Theory is so widely taken by academics to be verbal or numerical, that the very idea of visual theory seems puzzling. Anthropologist and film maker David MacDougall anticipated this when, in envisaging a shift from verbal to visual anthropology, he noted that 'the epistemological and methodological implications of such a shift are substantial' (1997: 292). On the one hand, if we start with the verbal theory, this tends to disempower the photographic image. On the other, if we start with the visual, photos (even in juxtaposition or series), typography, layout, and other forms of visual display are not good instruments for arguing with; so arguing and theory, pace MacDougall, may have to be replaced by different kinds of understanding. Sebald (2002) has demonstrated that many more questions can be asked about the relationship of words to photos, if photos are interleaved very precisely and captionless within the lines of a piece of writing. This is another story, and one well worth pursuing.[17] For, given the inbuilt tendency for contemporary social science theory either to chase the image away or to be incapacitated by it, we should start looking elsewhere for ideas about how to keep the balance of autonomy between photographic image and words.


1 See Chaplin (2002) which gives a detailed account of this project.

2 Others include 1) The Haverhill Project , by Chris Dorley-Brown, in which 2000 inhabitants of Haverhill, Suffolk were photographed and their faces morphed into one image to form 'The face of Haverhill in 2000'. And 2) The Staunton Millennium Book, which consists of photographs of all the inhabitants of this Herefordshire village in 2000; each photograph accompanied by a short account, supplied by the photographees themselves - of their names, occupations and interests.

3 The down side of this is that the frozen moment may cause other related moments to be distorted or pushed out of mind.

4 as part of a contemporary response to Goffman's 1979 method of categorizing family photos (see this paper, p.6.)

5 Rose Greenway at no 26 was troubled by the photo of herself. She said she did not like the expression on her face. She did not want a copy of the photo. Why she was troubled was that she assumed it to be evidence - and her idea of her own facial expression did not accord with that evidence.

6 See Bennett,1998, p.45, for a reference to Carolyn Steedman's account of the content and influence of British school curricula 1955-75.

7 For example, The Guardian, 19.8.98, p7. '"Sorry " Sun puts wheelchair wife back in picture' is more or less self-explanatory. In another case, the Guardian's chief photographer was caught out- having doctored a picture of Gordon Brown holding up the new red Budget box on 3.7.97. On 7.7.97 he was obliged to explain himself in an article in the Media section of the Guardian called 'We're framed!', and subtitled: 'When the Guardian doctored a photograph of the chancellor last week all hell broke loose. Picture Editor Eamonn McCabe explains'.

8 Information supplied by Jessops, Croydon on 8.7.04.

9 More accurately, it is only indirectly humanly produced. Nevertheless, there is an automatic process at the heart of what it is to take a photograph.

10 We are often told that our world is becoming increasingly visual. This is true of the world outside academe: the emphasis on the photo as physical trace can be seen in the law courts, where, as shown in Figure 24, a photo can count as evidence. Likewise, the emphasis on the polysemy of the photographic image is evident on television, in newspapers and advertising hoardings, where spoken or written words are always on hand to give meaning to the images.

11 Though some visual sociologists, members of the IVSA, do use the camera to provide evidence to support an argument.

12 Peirce (1931-58) in contrast to Barthes (1968) - though also concerned to theorise the meaning of the image - emphasised that the image should be given sufficient semantic autonomy, and according it its own mode of signification - the iconic.

13 There is a whole area of sociology - visual sociology -which deploys photographs as well as words. This visual sociology was, and in many cases still is, based on the idea that photographic realism is free of theoretical problems. Visual Sociology's eponymous house journal was renamed 'Visual Studies' in April 2002 (a sign of the times?).

14 Constructivist' as opposed to 'constructionist'. Constructivist art had its origins in pre-revolutionary Russia, and also in Holland. There are still Constructivist artists practising in Britain today.

15 See, for example, Gilbert G.N, and Mulkay M, 1984, pp.150-1.

16 See, for example, Hughes (1989)

17 I start to pursue it in part 2 of this paper.


BARTHES, Roland (1999a) 'Rhetoric of the image', in Evans, J. and Hall. S, (eds), Visual Culture, London Sage.

BENNETT, Tony (1998) Culture, A Reformer's Science , London Sage.

CHAPLIN, Elizabeth (2002) 'Photographs in Social Research: the Residents of South London Road' Everyday Cultures Working Papers (2),

FOUCAULT, Michel (1973a) 'Preface' from The Birth of the Clinic, London, Tavistock.

GILBERT, G. Nigel and Mulkay, Michael (1984) Opening Pandora's Box, Cambridge University Press.

GOFFMAN, Erving (1979) Gender Advertisements, London, MacMillan.

HALL, Stuart (1973) 'The Determination of News Photographs' in S. Cohen & J. Young, (eds.), The Manufacture of News. Social Problems, Deviance and the Mass Media, London Constable.

HALL, Stuart (ed.) (1997) Representation: Cultural Representations and Signifying Practices, London, Thousand Oaks, New Delhi: Sage/Open University.

HAMILTON,Peter (1999) 'Photographic Representations and National Identities' in Representing the Nation, a study guide, Milton Keynes, Open University.

HUGHES, Malcolm (1989) Painting Space, Gallery Space , London, Annely Juda Fine Art.

MACDOUGALL, David (1997) 'The Visual in Anthropology' in Banks M, and Morphy H, Rethinking Visual Anthropology, Yale University Press.

SEBALD, W.G. (2002) The Emigrants, London, Vintage.

SMITH, Greg (1996) 'Gender Advertisements revisited: A visual sociology classic?' Electronic Journal of Sociology 2 (1), January. < >