The SAGE Handbook of Visual Research Methods

by Luc Pauwels (editor), Dawn Mannay (Editor)

The second, thoroughly revised and expanded, edition of  The SAGE Handbook of Visual Research Method s  presents a wide-ranging exploration and overview of the field today. As in its first edition, the Handbook does not aim to present a consistent view or voice, but rather to exemplify diversity and contradictions in perspectives and techniques. The selection of chapters from the first edition have been fully updated to reflect current developments. New chapters to the second edition cover key topics including picture-sorting techniques, creative methods using artefacts, visual framing analysis, therapeutic uses of images, and various emerging digital technologies and online practices. At the core of all contributions are theoretical and methodological debates about the meanings and study of the visual, presented in vibrant accounts of research design, analytical techniques, fieldwork encounters and data presentation. This handbook presents a unique survey of the discipline that will be essential reading for scholars and students across the social and behavioural sciences, arts and humanities, and far beyond these disciplinary boundaries. The Handbook is organized into seven main PART 1: FRAMING THE FIELD OF VISUAL RESEARCH
PART 2: VISUAL AND SPATIAL DATA PRODUCTION METHODS AND TECHNOLOGIES
PART 3: PARTICIPATORY AND SUBJECT-CENTERED APPROACHES
PART 4: ANALYTICAL FRAMEWORKS AND PERSPECTIVES 
PART 5: MULTIMODAL AND MULTISENSORIAL RESEARCH
PART 6: RESEARCHING ONLINE PRACTICES
PART 7: COMMUNICATING THE FORMATS AND CONCERNS  

Genres: Research

776 pages, Hardcover

First published August 9, 2011

Reviews

[Trevor · Rating 5 out of 5]

There is an awful lot to this book, I haven't read it all (but probably about a quarter of the papers) and I'm really reading them with a kind of singleness of purpose - to help inform how I'm going to go about analysing photographs and other visual data. So, this review is partial at best Let me start by saying that this really is a wonderful book. The visual seems like it ought to be easier to talk about than the linguistic. To talk about words you really need to use other words - so it would seem that there is a built in contradiction here that ought to be pretty hard to overcome. Whereas, talking about images ought to be really quite simple - language seems to have come about as a way of talking about things and so you might think it was the perfect tool for the job of talking about images. The irony, then, is that so often we just don't have any tools for talking about images, and we tend to use work done in various schools of linguistics as a way of structuring the little we find we are able to say. There is a chapter in this book by van Leeuwen which in many ways gives a summary of his Reading Images I've reviewed previously. It goes somewhat further than this, too.

One of the particularly interesting articles is by one of the editors, a man I met at a recent conference here in Melbourne, Eric Margolis. His chapter is co-written with Rowe on methods of disclosing historical photographs is really interesting. The key problem here is that we generally see photographs as unproblematic windows onto the past. That they are not really affected by social tastes and preferences, but are just a way of freezing time - or as the authors say, they are like a mirror with a memory. Except this is far from being the case. A photographer not only gets to choose what they take, but also what they cut out of frame. They mention the work of Curtis who took lots of images of Native Americans and systematically ensured he cropped these images to remove any references to modern life. This kind of selective vision, one that refuses to notice and to actively hide certain facts is essentially ideological - but such an ideology is very hard for the viewer to see. Like I said, photographs have a verisimilitude to them and this makes it very hard for us to see they may not tell the whole truth.

They also discuss the limitations of content analysis of photographs. Content analysis is a kind of 'objective' analysis of what appears in photographs. There are so many problems here. One is to make sure you have a representative sample. Elsewhere Margolis has done an analysis of historical school photographs that have ended up on online databases. These tend to be very White and so the Black and Latino schools which existed in reality don't exist in this idealised world. Silences and gaps, then, become just as important as what is said and what is on show.

All the same, photographs - even the most staged - can give a glimpse into other worlds the photographer may have preferred to remain hidden. As is said elsewhere in this book, a photograph is different from a drawing as a drawing only shows what the artist 'noticed', but a photograph shows what is in front of the camera - whether the photographer noticed it when pushing the shutter button or not.

Despite the limitations of content analysis, I really do think it has its place. Quantitative Content Analysis of the Visual by Bock, Isermann and Knieper gives a really good introduction to content analysis and some of the problems people face in doing this style of analysis. This is very much a positivist approach to analysing images - the key things to worry about are how good a sample of images you have gotten and if the codes you have come up with could allow someone else to look at the photographs and get the same results you have. But these are hardly unreasonable expectations.

The authors say, "In general, a quantitative content analysis consists of five phases. First, the main unit (basic population) needs to be defined. In this phase, the researcher also decides whether to take a sample unit or to examine the whole research material and selects the unit of analysis (full pages, articles, pictures, info-graphs, and so forth). Second, the researcher compiles a code book, which contains the categories to be used to code the research material. This is followed by encoding in the third step and later the data analysis. Finally, the results are summarized and published. In the following section, we direct explicit attention to the issues of visual content analysis, which may occur in the first four phases."

As mentioned before this form of analysis seeks to be exhaustive and exclusive - you want to be looking at categories that sum up the pictures and you don't want similar images ending up in different categories because these categories are poorly defined. The problem here is once again that images don't just have one meaning. Think of a photograph of North Koreans weeping in the streets following the death of their great or dear leader. Presumably these images mean something quite different to the members of the Central Committee of the North Korean Party then they do to me - otherwise there would be no way they would let them be taken. Coding then has to be about things in the photograph, not about such reactions.

So, to understand a photograph you really need to understand the context in which it was taken and to do this you have to have some kind of theory - preferably sociological and stated up front rather than just assumed.

There is also some of those awful truths mentioned along the way, like this one, "Cunningham, for instance, demonstrated that recipients suppose attractive persons to be more intelligent and self-confident significantly more often than less attractive persons."

One of the really interesting things they say is that with press photographs we can assume that even he most 'natural' looking image of our leaders is probably staged. They show members of the G8 chatting away to each other, but rather than this being 'natural' if you look closely you notice each leader is sitting under a flag of their country. They have been very much placed in this scene and where they have been placed has meaning to how important they are in relation to the host nation. The closer you get to sit to the host leader the more important you are. Also all of the photographers are shown clumped together so that they are all forced to take virtually the identical shot of these leaders. The images produced are all perfectly in focus and well exposed. Now, this says something else that is really interesting, but that we might not know about ourselves. There is a sense that when we see a 'perfect' picture like this that we might not completely believe it. At least, we might not believe it in quite the way we believe a slightly out of focus image - one that has been 'taken on the run'.

I'm going to stop after this last one, but this was a particularly interesting article. Visual Semiotics - Key features and an application to picture ads by Winfried Noth. He starts by making the point that language is much better at dealing with temporal information - words develop along a narrative and a narrative is temporal in nature. And that images are better at the spatial. This isn't always true - there are no absolutes here - but it provides a good rule of thumb.

While you would normally have to do some pretty heavy footwork to establish that images are signs, in the linguistic sense, this paper talks about ads, and I think it would he hard not to see the point of ads as providing a series of signs that are meant to be read and understood and acted on. Rather than focusing on Saussure's concept of signs he relies on that of Pierce, an American philosopher who seems to have come to much the same conclusions as Saussure at much the same time.

Pictures can be icons, indexes or symbols - an icon stands in for another object, an index points to the object and a symbol has a meaning that is literally read into the image - like a traffic light, what Saussure would refer to as the sign's arbitrary, but conventional meaning.

To understand images we need to understand the cultural context in which they were made. But it is more than this too. Images have a syntax that is related to how we literally operate in the world, whereas language can be quite different. I need to make that more clear - in some languages the person who is doing the action comes first, in others it is what they are acting on that comes first, in other languages the verb, what is being done, comes at the very end of the sentence, in others in the middle of the sentence. All of this is up to the language. But we can't read a photograph as if rocks floated in the sky or the sky was down and the ground up. It is not that there is no semantics to photographs, and in a sense the semantics that are there are much more immediate than they are in language.

This really is a remarkable book - the more I learn about how images work on us and how we work on images the more fascinating I find this stuff.

[Катерина Майковська · Rating 3 out of 5]

There is a tone of valuable information in this book, however, I would not call it a handbook, but rather a collection of essays. I think such framing would help the reader better understand what they are getting into. Authors mostly discuss methods and sometimes provide interesting insights about the developments and disagreements in their fields (like in the semiotics chapter), and rarely provide new methods guidelines. I especially appreciated the chapters on Rephotography and on Maps. The first one on maps sparked an intriguing discussion about colonial nature of maps and what implications it might have on our research, but unfortunately, did not go much in depth.

[Nthoks]

This is very insightful piece