Social distance portrayed: television news in Japan and the UK

The potential of the camera framing, or shot-size, semiotic resource to encode meanings related to social distance has been recognized for some time. This study seeks to bring this resource into the remit of systematic analysis. Data are taken from screen measurements of portrayals of social actors in news programming produced by two national broadcasters, NHK in Japan and the BBC in the UK. Results for these two media outlets are compared and an attempt made to place the results in a meaningful cultural context. Analysis focuses on NHK’s images and the less familiar Japanese media system.


A B S T R A C T
The potential of the camera framing, or shot-size, semiotic resource to encode meanings related to social distance has been recognized for some time. This study seeks to bring this resource into the remit of systematic analysis. Data are taken from screen measurements of portrayals of social actors in news programming produced by two national broadcasters, NHK in Japan and the BBC in the UK. Results for these two media outlets are compared and an attempt made to place the results in a meaningful cultural context. Analysis focuses on NHK's images and the less familiar Japanese media system.

K E Y W O R D S
camera framing • Japan • shot-size • television news • UK

I N T R O D U C T I O N
Standards for social distance vary from culture to culture. If 'realistic' television programming such as news is doing what it claims to do, then these differences will be reproduced in the representation of social distance observable materialized in the 'camera framing' semiotic resource. Variations across broadcasters can cast light on the relationships between broadcasters, subjects and viewers.
This article compares news images from two countries, the UK and Japan. The broadcasters chosen to act as source for the images are national public service broadcasters NHK and the BBC. These two organizations are 'related' in that they both operate in similar regulatory environments, the BBC being governed by its charter and NHK by the 1950 Broadcast Law; they both operate within a dual public-service/commercial environment and they are both funded by audiences, the BBC by licence fee payments, NHK using a mechanism of paid receiving contracts. 1 These broadcasters produce news programming, visual and linguistic material in the form of V i s u a l C o m m u n i c a t i o n 1 2 ( 1 ) news 'stories', for their audiences, and by doing so offer a view of the reality of the external world.
However, the UK is not Japan and vice versa; NHK operates within a Japanese cultural context and the BBC within that of the UK. NHK programming is created by a majority Japanese staff and for a majority Japanese audience. Likewise, the BBC tailors its output to a UK audience. How these broadcasting cousins choose to portray social actors to their audiences can shed light on conceptions of social relations within the respective societies.
Images are approached here as the embodiments, the material result, of the activities of socially-embedded individuals. The image is a text open to analysis in terms of the 'semiotic resources' (Kress and Van Leeuwen, 2006) mobilized in semiosis. My focus is on camera framing, or shot size, as one particular semiotic resource, because of its ubiquity, its measurability and the important role it has been theorized as playing in portrayals of 'social distance' (Hall, 1996(Hall, [1990). Preferred social distance has been shown to vary across cultures (e.g. Beaulieu, 2004;Sussman and Rosenfeld, 1982), the individuals involved in media production in a particular culture are generally part of that culture and might be expected to reproduce social understandings of that culture in their created texts.
Presented here is a quantitative comparison of representations of social distance in the news coverage of two public service broadcasters operating within two different mature media systems. The UK and Japan, while having much in common as developed post-industrial societies, have emerged in different geographic and historical contexts; the UK being part of the European Judaeo-Christian cultural milieu and Japan as part of the East-Asian Sinocentric Buddhist/Confucianist region (Brown, 1998: 18-32). The press systems they support have thus grown in different cultural soils and can be expected to show variation: National and regional philosophies form the foundations about the press in countries throughout the world. A certain set of ideas about the relationship between the press and a society derives from longstanding beliefs and intellectual traditions. (Winfield et al., 2000: 323-324) Such culture-born differences may be attenuated by their materialization within the essentially universal physical-technological framework of television production. This study looks outside the text to the context of production and draws in understandings of the surrounding technological, cultural and social environment. It should be mentioned from the start that, whilst comparative, this study uses the BBC primarily as a point of reference to discuss the images of NHK's news.
A method is proposed for producing quantitative data which describe representations of social distance on television by taking facial measurements of images sampled from news programming. The process is straightforward and reproducible, all data so derived will be comparable, thus allowing a broader scope of images to be subjected to systematic examination. This will benefit the project of de-westernisation of media/semiotic theory and be of value to those involved in comparative media studies. Derived data describe variations in shot-size used in NHK and BBC news, and show the BBC preference for wider shots and NHK's use of a greater proportion of closer-up shots. These findings are analysed in the light of material drawing on the broader context mentioned above.
By focusing narrowly on the mobilization of one particular semiotic resource in different media systems, we can gain insight into both the role of the resource in the media systems concerned and the efficacy of theory concerned with describing the resource. Theories I make use of have been developed largely in a western cultural context; it is healthy, and indeed necessary, that they should be tested against a variety of data.

T H E O R E T I C A L B A C K G R O U N D : S O C I A L D I S T A N C E A N D I M A G E S
A social semiotic approach: This study looks at how certain aspects of social activity -what becomes 'news' -are manifested in the visual texts of television news and what assumptions about the world are implied by the mutual comprehensibility of these texts. Producers construct these texts in order (we assume) to communicate, viewers view them and in the majority of cases are able to understand and interpret them adequately. Given the social/media environment in which the texts examined here are circulated -one in which elementary feedback in the form of ratings is readily available -it is unlikely that a producer who continually presented texts which were incomprehensible to the greater part of their audience would be allowed to continue doing so for long.
As a general approach to the exchange of signs which constitutes this communication, I adopt a 'social semiotic' understanding of meaning-making as rooted in social practice and in which 'texts ... are defined as being the semiotic manifestation of material social processes' (Hodge and Kress, 1988: 137). This allows a broader, socially engaged analysis of meaning-making which is to be preferred over the more formalistic and context-independent view of semiosis implied by 'general semiotics' , especially when considering the predominantly non-authorial products of mass media.
This study looks at framing size as one instance of a semiotic resource, described by Van Leeuwen (2005: 4) as follows: Resources are signifiers, observable actions and objects that have been drawn into the domain of social communication and that have a theoretical semiotic potential constituted by all of their past uses and all their potential uses and an actual semiotic potential constituted by those past uses that are known and considered relevant by the users of the resource. Camera framing -I use this phrase to distinguish my subject from 'framing' in the sense of an organizing concept in journalistic activity (e.g. Gitlin, 1980;Tuchman, 1978) -or shot size (I use the two interchangeably) qualifies as such due to its pervasiveness and the consideration it is given by media and film theorists, those involved in practical media training and camera operators themselves. Indeed, one of the fundamental and indeed definitional differences between 'pictures' , whether paintings, photographs, films or television, and the world as we experience it through our own unmediated sense of vision is the fact that the former are 'delimited'; they have an edge, they start and end somewhere. Gibson (1971: 31) offers this definition of a picture as a surface so treated that a delimited optic array to a point of observation is made available that contains the same kind of information that is found in the ambient optic arrays of an ordinary environment. (emphasis added) The decision regarding where this delimitation occurs is, in the first instance, that of the image producer. As far as videography is concerned, the general shape of the frame, the limits of the chosen segment of visual reality to be 'pictured' , is the result of technological convention. The camera operator can decide, within the limits of the technology of lenses and lighting available at the specific moment of creation, where to place the edges of the picture, what to include and what to exclude. When this decision is taken it is impossible, with the relatively sparse resources and within the time constraints available to news production organizations, to add anything back into this picture, to reinstate what has been excluded. Deletions, contractions and cropping of the image, on the other hand, are much more easily achieved and, indeed, one type of cropping, that allowing for variations in the screen ratio of domestic television receivers, is so prevalent at the time of writing that allowances for it to take place are part of the typical process of image production.
The various choices made by the image producer, the ways different shot sizes are used in the portrayal of social actors, have direct implications for the way images are read and understood. They affect how close or distant we seem to be when viewing the image, they also embody a certain degree and nature of physical, psychological and, ultimately, social involvement encoded by the image producer. The following section describes the linkage between framing sizes and social distance as currently theorized.

The hidden dimension
ET Hall coined the term 'proxemics' to describe the study of 'social and personal space and man's perception of it' . His 1966[1990] work, The Hidden Dimension, in which he elaborates his theory, deals with spacing in animals and humans, how perceptions of space and spacing are expressed in language and visual art and specifically with conceptions and expressions of appropriate interpersonal distance in three European countries, 'the Arab World' and Japan.
The crux of Hall's work is his theory of 'distances in man' and it is with this, and its relation to the size of the field of human vision and the proposed connection with certain 'standard' image sizes used in film and television, that the following paragraphs are concerned.
Different 'social distances' imply different fields of vision: the closer one is to someone, the less one can see of what surrounds them and the more detail of their features. Hall's degrees of 'social distance' (see Table 1), it has been suggested, bear some relationship to the typical framing sizes used during image production (discussed in section 1.2); simply put, close-ups may indicate intimacy and a subject shot from a 'wide-shot' distance may seem just that, distant (Hall, 1996(Hall, [1990). Hall (1996Hall ( [1990) identifies four ranges of social distance -intimate, personal, social and public -divided into 'far' and 'close' subdivisions, referred to as phases, they are described as follows: Intimate distance: 'This is the distance of love-making and wrestling, comforting and protecting. Physical contact or the high possibility of physical involvement is uppermost in the awareness of both persons' (p. 117). In the far phase 'the iris of the other person's eye seen at about six to nine inches is enlarged to more than life-size ... Clear vision (15 degrees) includes the upper or lower portions of the face' (p. 117).
Personal distance: 'The distance consistently separating the members of a non-contact species' (p. 119, after Swiss animal psychologist Hediger). Hall refers again to the physical possibilities of contact implied: 'at this distance one can hold or grasp the other person' (p. 119); it is to be expected that the psychological impact of a reproduced image of a subject experienced from this distance will be reduced in comparison to that of the presence of the subject him/herself. Social distance: At which 'intimate visual detail in the face is not perceived, and nobody touches or expects to touch another person unless there is some special effort … Impersonal business occurs at this distance' (p. 121).
Public distance: 'Clear vision' can take in the faces of two people at twelve feet and 'scanning vision' can include the whole body, individuals at this distance are -if they wish to be -'well outside the circle of involvement' . 'Fine details of the skin and eyes are no longer visible. At sixteen feet, the body begins to lose its roundness and to look flat ' (pp. 123-124).
Social distance in Japan and the UK: Hall's comments on Japan are restricted to the conceptions of space in giving and following directions and the concept of 'interval' (ma) in traditional architecture and garden design, his only comment on possible differences, compared to the base of his comparisons, the US, is to state that 'the Japanese I have known prefer crowding, at least in certain situations' (p. 152).
Other works, unfortunately few, have studied interpersonal space and conversational distance in a comparative context that includes the UK and Japan. Social psychologists Sussman and Rosenfeld (1982: 70) suggest that a typical interpersonal distance for Japanese-speaking dyads engaged in casual conversation is in the region of 1m, greater than that adopted by North Americans speaking English (c. 90 cm) and Venezuelans speaking Spanish (c. 80 cm). However Beaulieu (2004: 800) suggests that Anglo-Saxons (people from the UK, and English-speaking North America) in fact adopt greater mean personal distances, in the region of 80 cm, than do Asians, including Japanese, at about 70 cm. Typical interpersonal distances vary between the UK and Japan, this study hypothesizes that the television news media will reflect this in its portrayal of social actors.
Subjective perception of reality -what interpersonal distance is just right -is only one possible factor in the creation of representations of social situations for television news, other influences need to be identified and discussed. This study's analysis attempts to take in this broader array of factors while concentrating on the central notion of social distance (see section 5). Kress and Van Leeuwen (2006: 124-129) suggest that the 'standard' shot sizes used in film and television relate closely to the different fields of vision implied by Hall's degrees of social distance. Ferguson and Ferguson (1978: 27) also suggest a correspondence as follows:

Social distance and the televisual image
If the communicator is shown in a close-up shot of the head only, this would correspond to a physical intimate space arrangement. A medium or waist shot would correspond to personal distance. A medium long shot showing the full subject would correspond to social distance; and a long shot, where the subject comprises a relatively Figure 1 shows seven images which recreate, using 3-D modelling software, 2 the view of a human figure from the boundary distances of Hall's categories of social distance. The figure is 1.80 m tall and observer eye-line is set at 1.65 m above floor level, observer 'field of vision' is set to 35° to take in the beginnings of peripheral vision which Hall defines as starting at around 30°.
The match between the images of Figure 1 and the descriptions of some of the standard shot-sizes in Thompson (1998, see Table 2) is remarkable.
These shot-sizes, and the social distances they imply between the portrayed and the viewer, further imply certain physical, psychological and social relationships. The image producer, in the case of television news, the camera operator who chooses where to delimit the ambient optical array, is therefore intimately involved in the process of constructing the relationship between viewer and portrayed participant.
The implication of physical distance expressed through the resource of framing size is highly realistic (I recognize that this term is problematic and I use it here in the sense of highly transparent or motivated, that is needing little if any conscious interpretation) and is relatively straightforward -the larger an object appears, the closer it seems to be. This is in contrast to interpretations of this distance within a specific social context, which are more nuanced. However, the immediate and instinctive recognition of the physical distance component of social distance implied by the framing size of a human image somewhat limits the bounds of possible 'expert' insight. When it comes to judging how far we are from somebody and what sort, very generally, of relationship that might entail, we are all experts and our instinctive interpretation is generally good enough for most situations we encounter.
Below are some examples of the types of interpretations placed on the uses of frame-size as a constructor of social distance; in broad terms, the closer we approach another individual, the more we become entangled with him or her socially, psychologically and, perhaps, physically. As Hall (1996Hall ( [1990) comments, intimate distance is the distance of 'love-making and wrestling', the necessary closeness to engage in direct physical contact, whether that be a caress or a blow. Hodge and Kress (1988: 53) characterize these relations as 'strong', whereas 'non-closeness normally signifies weakness, indifference, or alienation in a relationship, either positive or negative'. It is further suggested that both remoteness and closeness are ambiguous in term of their signification, with closeness being more open to interpretation than remoteness. This would seem to be common sense as there are more ways to interact with an individual if one is in physical proximity: at close range one could tickle, pinch, breathe on or even lick another person; when people are outside the range of direct contact, they are reduced to using one of the 'ranged senses', they can see and expect to have their gestures seen, or can use their voice and expect to be heard. It follows that the greater range of interactions possible at close approach will imply a more ambiguous interpretation of portrayed 'closeness'.  Thompson (1998 In his instructional work Videojournalism, Griffiths (1998: 95) offers one particularly concrete interpretation of the extreme close-up: 'If your intention is to make somebody appear shifty and untrustworthy then by all means take an XCU.' Articulation points and edges: Frame sizes, while apparently related to Hall's social distances, are also influenced by human physiology, the inevitable act of delimitation entailed by image production, and by what can be described as a visceral fear of dismemberment. Common frame sizes avoid the edge of the frame intersecting with any of the obvious points of articulation of the human body, knees, the neck, etc. for the simple reason that seeing a body so framed, it is difficult to avoid the impression, however unreasonable or unlikely we know it to be, that the edge of the frame, the camera, has somehow lopped off the unseen extremity (Musburger, 2002: 126;Zettl, 1973: 159-160). Figure 2 shows an example of such an image.
Summary: Camera framing here then is assumed to be a semiotic resource which can be mobilized by image producers to express, or encode into their produced texts, messages whose meaning is concerned with social distance. The cultural experience of social distance is a component of the reality that news producers seek to 'objectively' report.
The following section deals with the problems posed by the concept of camera framing sizes and describes the terminological mire which the methodology proposed by this study attempts to bypass.

C A M E R A F R A M I N G : P R O B L E M S A S A P R A C T I C A L C O N C E P T
Images produced for television conform to the shape of the destination medium, the television screen. It is thus necessary to decide, in relation to the scene encountered by the camera operator, where to set the edges of the section of the scene to be recorded. The camera operator must choose to frame Figure 2 The subject here has been shot from the neck down in order to maintain his or her anonymity. In the absence of any such necessity, placing the edge of the image at a natural joint such as the neck is avoided, for obvious reasons. It is difficult to avoid the phrase 'head chopped off' somewhere in the mental processing of this unusual image. the scene within the technological limits inherent in the technology of televisual image gathering, processing, distribution and viewing. In theory at least, each image can therefore be described in terms of the extent of this framing in relation to the scene or object of interest.
The most common way of describing framing sizes is through a set of terms specifically created for that purpose; these are familiar to anybody in the film or television production world and have also in some cases, for example 'wide-shot' or 'close-up' , become part of general speech. However, one encounters problems when attempting to use this terminology in a scientific context. The following section outlines these problems and describes the methodology adopted to derive the data used in this study.

Approaches to framing size
Literature, whether academic or instructional, that deals with framing sizes, is inconsistent in both the descriptive terminology used to label shot-sizes and the sizes of images to which they refer. Given the origin of frame sizes in the 'craft-like' context of their use in television and film production, and the highly variable standard of their measure (the human body), this is hardly surprising.
Frame sizes are 'relative' descriptions of the relation of the subject to the extents of the visual image rather than absolute measures. For example, an MCU extends from just above the head to about mid-chest whether the subject is 1.5 m or 2 m in height, regardless of the amount of the individual actually portrayed measured in cms or inches. Figure 3 shows on the left a human figure measured in 'heads' and, to the right, six linear scales indicating the lower edge of framing sizes as described in the various sources listed. The upper edge is assumed to be just above the top of the head in all cases except for those framings (BCU, VCU, XCU, Up) which are less than one head in height, where the upper edge for these 'close-ups' is between the hairline and the eyebrows. As can be seen, while there is a degree of agreement, nothing here resembles a consistent measurement scheme. If images are to be discussed in a consistent fashion by researchers then this terminology has to be deemed unfit for purpose in its current state.

Summary of problems
As well as this terminological inconsistency, other obstacles stand in the way of the would-be analyst of camera framing. These can be summarized briefly as follows: Non-human subjects: The traditional shot-size descriptions are premised on the presence in shot of a human subject. They do not allow for consistent description of images of non-human subjects.
Size-posture-attitude: Illustrations that demonstrate frame sizes in production text books and academic works such as the ones mentioned above (see Figure 3) invariably show standing, sometimes walking but always upright, adult human figures, and frame sizes are related to whereabouts on such a figure the delimiting edge of the frame would fall for a particular type of shot. Portrayals of non-adults or individuals shown in non-standing postures are thus problematic.

Alternative approach
This study answers, or avoids these problems as follows: firstly, by adopting a method which uses measurement rather than classification to describe framing sizes; and, secondly, by adopting a comparative approach and arbitrarily eliminating all non-human and non-measurable portrayals from the offset.
The latter move might seem somewhat extreme but, given the objective here, it is essential to establish a rigorous basis on which comparisons can be made. Furthermore, such a move was deemed acceptable in view of the fact that the primary concern of both broadcasters' news is the action of adult humans.
So, instead of the far-from-standard descriptive terms, this article uses actual screen measurements, more accurately the ratio of facial length to screen height, to describe framing sizes. It offers an objective (as far as this term can be applicable when describing any account of an image that relies on human perception) analysis of news images from two countries' media systems. The following section details this methodology.

M E T H O D O L O G Y
Source material: This comprised edited video packages from NHK and BBC news programming, these being the most widely viewed news broadcasts in the two nations; the BBC's Six O'Clock News had in the region of five million viewers in Nov 2010, NHK's News 7 on 5 Nov 2010 gained ratings of 18.3 per cent, roughly equivalent to 18-20 million individual viewers nationally. 3 Sample selection: Sampling procedures were largely dictated by the practicalities of obtaining recordings. NHK recordings were made during summer 2007 (part of a broader study of news images in Japan, see Koga-Browes, 2009), BBC recordings were made in summer 2010. During the time between the two sets of recordings, the prevalence of wide-screen had become established, thus the earlier NHK recordings were in 4:3 and those of BBC programming in 16:9 wide-screen. This methodology uses screen height in its calculations so, despite this, the sets of images were deemed technically comparable.
Story selection criteria: All live material, studio-based and outside broadcast, was excluded from the sample; these segments, due to limitations of physical space and technology, tend to be visually simpler than edited packages with less scope for 'creative' decision making about image content and thus offering less potential insight into image producers' world-view. Likewise, all foreign news coverage was excluded as it may rely extensively on agency footage, the origins of which are unknown to the viewer/researcher. Stories shorter than 2 minutes and longer than 4 minutes were discarded -these cutoff points were chosen primarily in order to return a manageable amount of data whilst capturing the majority of packages.
Stories were sampled, a 'still' extracted, at 2-second intervals, this interval being chosen in order to capture an image from every cut, almost none of which are shorter than this This resulted in 4039 images from which measurements were taken of the proportion of screen height occupied by portrayed individuals' faces. Roughly 3 in 10 images portrayed no human participants, 2846 facial measurements were derived from the images. In order that comparisons were valid, a roughly equal number of data was derived from NHK and BBC packages: the 1422 pieces of BBC data came from 26 stories, the 1424 NHK data from 25 stories. NHK stories tended to be longer, thus yielding more samples (c. 86 on average, BBC: c. 72) but contained relatively fewer portrayals of social actors, 66 per cent of samples as opposed to 75 per cent for the BBC images. The corpus data are summarized in Table 3.
Portrayed social actors' face size was measured, base of chin to hairline, using on-screen measurement software. 4 Where several social actors were portrayed together, it was necessary to judge which was the focus of the image, this somewhat compromises the aims of the study -to engage with images in as objective a manner as possible. However, it is also advantageous in that, given that there may be more than one image sample for longer cuts, it allows measurements to be taken of more than one social actor per cut, if necessary. For the same reason, several measurements can often be taken of the same portrayed individual thus leading to more reliable overall results. In all cases, only one measurement per sample was made. All measurements were then converted into ratios dependent on the screen height, which varied from recording to recording.

Shortcomings and limitations:
This method is prone to many of the same limitations as a method which relies on classification of cuts into shot-sizes (see section 2.2). It can only deal with images of human beings, the identification of the subject of an image has to be performed by the analystopinions may differ as to what the image is an image of, the face of a portrayed subject may not be fully visible, or the subject may be portrayed in a posture, or from an angle, where foreshortening affects the accuracy of the measurement. However, it does go a long way to overcoming the problem of terminological inconsistency and can thus, I believe, be considered a step towards an empirical engagement with televisual portrayals of social actors.

R E S U L T S
The results data are summarized in Tables 4 and 5. As can be seen, NHK used three times more very close-up shots, those where the portrayed social actor's face more or less filled, or overflowed the screen (90 ->100% of screen height), than the BBC. And while these types of shots only make up just 1 per cent or so of the total, this tendency for NHK to use a greater proportion of closer shots is evident from Figure 4, which shows NHK outnumbering the BBC in all regions except one for shots that portray a face filling half the screen height or more. On the other hand, the BBC uses portrayals which tend to the distant end of the spectrum; in nearly 60 per cent of portrayals of social actors, the faces of those shown occupy less than 30 per cent of screen height, for NHK such representations make up under 40 per cent.
The 2846 measurements derived were also graphed for examination: firstly, according to arbitrary regular divisions (see Figure 4), each division representing a 10 per cent point change in the face-size:screen-size ratio; secondly, they were also considered in terms of the standard, traditional shot-size categories discussed in section 2.1. The resulting graph shows clearly how and where the two companies' portrayals of human individuals diverge.  It should be noted that, due to the method used, measurements greater than 1, where the screen height was taken up entirely by face, are estimates, either the forehead or chin or sometimes both not being visible; they are thus not considered in any detail here and are simply counted as 'greater than 1'.
Standard shot-sizes: In order to give relevance and cross-comparability to industry/practice-related discourses, the data were also analysed in terms of the standard shot-sizes mentioned above (results in Table 5 and Figure 5). Face measurements for these shot-sizes, that is the bins used to create the histogram, were based on the images derived from Hall (see Figure 1). These data representations all tell very much the same story. NHK uses a greater proportion of closer-up shots than does the BBC, using more than double the number of CUs while counts for other framing sizes are roughly similar. The BBC tends to prefer looser shots and generally outnumbers NHK in the MCU to LS range.

A N A L Y S I S A N D D I S C U S S I O N
Portrayal of social distance: The following section focuses on the elements that seem to distinguish NHK's images from those of the BBC. For this reason, it concentrates on the Japanese media system.  To summarize, each image can be considered to represent, perhaps rather roughly, two seconds in time. Thus, the news viewer in Japan spends almost double the amount of time viewing human faces which take up more than half of the screen hight than does the viewer in the UK; images for which the 'facial height' to 'screen height' ratio is greater than 0.5 number 196 for the BBC and 364 for NHK.
While it is difficult to characterize a typical shot-size for each of the two broadcasters, it can be seen that NHK prefers the CU over the MCU and the BBC prefers both MCU and MS over the CU. The average proportion of screen height filled by a face on NHK is 0.365, while for the BBC it is 0.286 -these sizes are illustrated in Figure 6. As can be seen in these images, the necessary by-product of using less screen-space for foreground material (here speakers' faces) is that more space is left to be filled by background. 5 On textual evidence alone, it is not possible to distinguish between close-up images created in order to maximize space given over to foreground material and those perhaps created to minimize background material. However, the outcome is that the closer-up shots preferred by NHK tend to reduce the amount of location-related information present in the background.
Translated into social distance terms, NHK's typical portrayal can be interpreted as portraying a closer relationship between viewer and viewed while the BBC is rather more 'distant' , perhaps stereotypically 'standoffish' . The social distance implied by the typical NHK image is in the region of 0.88 m; for the BBC image, this is approximately 1.13 m. 6 Putting these results in the context of the findings mentioned at the end of section 1.1, the picture that emerges is a confused one (see Table 6). The value for NHK falls more or less in the centre of the range described by Table 6 Summary of social distance findings data (cms).

UK/Anglo-saxon
Japan/Asia Sussman & Rosenfeld (1982) apx. 90 apx. 100 Beaulieu(2004) apx. 80 apx. 70 this study 113 88 the two other studies but the value for the BBC lies well outside. This could be interpreted as follows: • the results, some or all of them, are inaccurate, or • they actually measure different things, or • the results from this study reflect differing attitudes in the ways the broadcasters portray society.
The second and third possibilities are interrelated, I deal with them together after considering the first. Generalizability: The studies referred to at the end of section 1.1 have no obvious flaws. Of course, they can all be criticized for not using larger or different samples but, given the limitations of academic enquiry, they are sound. Having said that, both Sussman and Rosenfeld (1982) and Beaulieu (2004) rely on observation of subjects removed from their natural cultural surroundings -overseas students attending US universities and attendees at an international academic/industrial conference. These individuals are unlikely to be typical representatives of their national populations being in general young, highly educated and possibly members of an economically privileged elite. It may well therefore be difficult to extrapolate from these results to the society from which the subjects are drawn.
Direct vs mediated communication: Furthermore, it could be argued that the results derived from these studies of non-mediated communication, with its limited yet concrete sample of individuals, and the results derived here -based on a postulated relationship between two overlapping, shifting groups, 'the mass media' and 'the audience' -are not in any sense comparable. In which case, the second objection would be true.
The mediated communication of television production/viewing is not conversation, being structurally more complex. A distinction has to be made between the three possible vectors of social distance involved: firstly, there is the social distance between the broadcaster and the subject; secondly, that between the subject and the viewer; and, thirdly, that between the broadcaster and the subject constructed for the consumption of the viewer. For the producers of the image, there is the actual relationship between themselves and the subject, in most cases here a news source, and the version of this relationship they wish to project to the viewer. To distinguish between these on the basis of the produced texts is impossible and perhaps near impossible even with access to the image producers.
On top of this is the question of whether or not the two broadcasters are using the camera framing resource in the same manner. It may be that both broadcasters are not equally conscious of the way this resource plays a role in constructing relationships between producer, source and viewer. One or the other may be more pro-active in the use of camera framing to project its own self-image to its viewers. Unfortunately, such questions fall outside the limitations of this study.
What is possible: Given the preponderance of doubts that seem to accompany the results of this study it seems only wise to severely limit the conclusions drawn to those which have been demonstrated.
To restate, NHK, the national broadcaster of Japan, prefers to use comparatively fewer image of social actors, but the images it does use tend to present these social actors using tighter camera shots. On the other hand, the BBC, while using more images of social actors, typically uses wider camera framings in their presentation. The following sections look at factors which may account for some of this variation. Pharr and Krauss (1996: xi) describe Japan as the most 'media-saturated' country in the industrial world. Krauss (2000: 266) uses the same term in his analysis of the relationship between the news media and politics in Japan: a 'media-saturated democracy'; likewise, Painter (1996: 197) when drawing comparisons between Japan and the US.

Pressure from the media environment
What media-saturation consists of, when it comes about, how we can recognize and quantify it is not discussed -a definition would probably ask more questions than it answered. Saturation carries with it meanings of over-filling and over-supply, going beyond satisfaction, putting so much into something that it can absorb no more. What this might mean when linked to the term 'media' is a question I must sidestep here. Instead I will refer to known facts that illustrate the penetration of various media in Japan. In 2005, the average Kanto-area household had the television on for just over 8 hours (between 6:00 and midnight) per day, the average individual viewed just over 4 hours during the same period, 70 per cent of viewers nationwide had five or more free-to-air commercial channels available, as well as two NHK channels (MIC, 2007: 152;7 MPR, 2007: 364-365). In 2006, over 90 per cent of Japanese households owned some sort of mobile phone, many with internet access and 'one-seg' television, and over 80 per cent owned a PC; overall, 80 per cent of Japanese spent more than 30 minutes a day reading a newspaper (MIC, 2007: 13, 94).
These figures offer a glimpse of the depth of media penetration and use in Japan but to characterize the actual experience of media in daily life in one of Japan's cities -where the majority of people now live and work -is more difficult. City centres, especially the areas around train stations, are rich in linguistic, visual and aural information; in a short walk, one may encounter thousands of signs, notices, fliers, leaflets and stickers, all hoping for a moment of attention. On top of this may well be a layer of aural appeals, safety and commercial announcements from the station, shop workers with megaphones, estate agents playing endless loops of their advertising jingle. Whether one chooses to see this as urban vitality or sensory vandalism, the informational density of urban Japan is hard to deny. 8 In such a landscape, appeals for attention must be immediately captivating -the appellant must attempt to seize the attention as rapidly as possible or risk its straying elsewhere. It is hardly surprising that these appeals should be made by attempting to engage the emotions rather than the intellect. Shock, amazement, surprise grab the attention more immediately than dry, factual explanation ever could.
NHK news, however minimally, may have become involved in this clamour of emotional appeals. The preference for 'closer' images of social actors may be an attempt to increase their emotional impact on the viewer, to make the stories they tell more dramatic, more immediately gripping and to thereby better hold viewer attention. Hardy et al. (2010: 18) suggest that, at the national level, 'more competition means more sensational news' and competition between broadcasters, and between the broadcast industry and other forms of information/entertainment, can certainly be said to exist in Japan, although how one might quantify this is a moot point. If so, then NHK, virtually a synonym for 'staid' , can probably be seen to be undergoing the beginnings of the process of 'tabloidization' attested to by studies of media around the world. 9 Having said this, the fact that NHK uses comparatively fewer images of social actors would seem to argue in the opposite direction. The tendency to 'personalize' news, to concentrate on concrete actors rather than processes or institutions, has also been identified with sensationalist styles of coverage (Bennett, 2002).
Visual tabloidization: The use of close-up images in an attempt to heighten the emotional impact of a news story is a standard component of the definition of the process of news 'tabloidization'; Hjarvard (2000) refers to the type of coverage offered by Denmark's tabloid news television as 'proximity news' , 10 highlighting its reliance on the breaking down of the barrier of distance/separation between portrayed social actors and news viewers. Hjarvard implies that one feature of the visual presentation of such news is the use of 'ultra close-ups'; although the term is not defined and illustration is not provided, it seems safe to assume that this is something like the VCU or XCU discussed above, where the portrayed's face more than fills the screen, shots generally focusing on the mouth or eyes. The term 'underdistanced' has also been used to describe sensational news stories in that 'they violate a comfortable psychological distance between audience members and their perceptions of events in the physical world' (Grabe et al., 2001: 637). Grabe et al. (2000Grabe et al. ( , 2001 also identify the zoom-in camera movement as a component of sensationalism and, while close-up images are not mentioned as such, the logical result of the zoom-in is often a close-up. NHK may have been caught up in this expressional arms race, in competition as it is with four, in some areas five, commercial networks whose news products have consistently made more use of coverage focused on individuals rather than events or processes, and the visual and sound effects theoretically associated with 'tabloidization' or 'sensationalism' in news (Hagiwara et al., 2001: 88-98), and against whom it competes for the ratings needed to justify its special status and its legal right to collect viewers' licence fees.
Ultimately, when considering the causes behind the observed results, it is almost impossible to differentiate between those close-ups which might reasonably be put down to leanings towards sensationalism and those which are the result of a desire on the part of the image-maker to portray the subject at what feels to them like an appropriate social distance.
On-screen linguistic material: Since the 1980s, the pictorial elements of the news image have had to share screen-space with an increasing amount of linguistic material (Kawabata, 2006: 210), in the shape of captions, bugs, date and identifier straps, and so on. NHK devotes more screen-space to these linguistic elements than does the BBC (see Figure 7). The corollary of this tendency to reduce the amount of space available for pictorial elements would seem to be a preference for wider shots; these leave more space around the main visual content, yet the opposite is true -NHK prefers tighter shots. Any investigation of why this is so would require a significant expansion of theoretical purview and must be reserved for another occasion.

C O N C L U S I O N S
Variations in certain social norms, in this case the typical physical separation adopted by members of a culture, are reproduced in realistic portrayals of that society, such as we assume news programming to be.
This study argues that, by taking measurement of portrayals of social actors from different broadcasters' news programming, it is possible to develop an index which represents the typical social distance between pictured social actors and viewers portrayed by that broadcaster and thus provides an insight Space occupied by linguistic extras in BBC and NHK news programmes. Areas outlined in black show the size of the 4:3 screen, the additional grey-outlined areas the 16:9 screen size. Area 1 is used by NHK to display the name, title, affiliation, etc. of a speaker, the BBC uses the darker grey area marked z for the same, the small additional tab under the left hand end of the main strap displays a BBC News logo only when the caption is visible. For NHK area 2 is used to display a headline for the story and area 3 a transcription of the speaker's words.
into the nature of the relationship, linking subjects, viewers and the broadcaster, envisioned by that broadcaster.
The methodology summarized here attempts to minimize the intervention of subjective judgements by the observer. It relies on actual measurement rather than estimations and classification of images into arbitrary and often inconsistent categories. The straightforward methodology is open to all who would seek to perform research on televisual images. This makes the area of quantitative-based research into visual expression open to a broader spectrum of participants and may lead to insights into more, and more varied, types and sources of images -something that can only enrich our understanding of human visual expression. A broader and methodologically consistent treatment of visual material which results in data sets that can be meaningfully compared should also lead to advances in theoretical understanding.
Theoretical problems and limitations: Further to the methodological problems already outlined (see section 3), there are more fundamental theoretical problems which arise in the context of comparative work such as this. Namely, the underlying assumptions that: • shot-size is meaningful, that it acts as a semiotic resource, and • that this semiotic resource is mobilized with similar intent in both cultures.
Until such problems can be resolved satisfactorily, analysis must remain speculative. However, what can be said with certainty is that the choices made by image producers about what kinds of portrayal need to be created and distributed as news lead to outcomes which vary in a quantifiable manner. That is, while the differences may be glossed as style, or feel -unquantifiable in any real sense -this study allows us to point with clarity to a manifest and documentable difference between the portrayal of social actors in television news in Japan and the UK. As such, it contributes to understanding how terms like those mentioned above might be usefully unpacked for analysis. Hall (1996Hall ( [1990: 113-116) seems to be ambivalent towards cultural variation in absolute social distances; his statement that people have a standard way of handling distances from their fellows and his emphasis on the physical and physiological origin of social distance -'the regularity of distances observed for humans is the consequence of sensory shifts' -seem to be arguments for species-wide consistency. However, Hall acknowledges that his work is 'a first approximation' and that it may not necessarily be possible to extrapolate even beyond his subject group of adult, white, middle-class, northeast-American 'intellectuals' . In the end, he seems to argue for the universal existence of the various phases of social distance he describes whilst stating that other groups have 'very different proxemic patterns' . This variation between groups (whether we should properly categorize them as national, ethnic, cultural or otherwise is a question beyond the scope of this article) has been confirmed by studies such as Sussman and Rosenfeld (1982) and Beaulieu (2004).
Human beings are socially and culturally conditioned, they are also physically embodied, social distance would seem to be the result of interplay between these physical and nonphysical elements. The physical nature of the body sets inner and outer limits, related to the activity and acuity of the various senses, to where the various transition points may occur; individuals learn from the social milieu in which they are raised which particular distance, within these limits, has been settled on as 'comfortable' .
It may then be the case that the physical distances arrived at in this study should be considered as marking the same social distance. That is, perhaps, rather than saying NHK prefers portrayals somewhere towards the 'personal close' end of the personal division and the BBC prefers portrayals nearer the limit of the 'personal far' subdivision (see Figure 8), it may be more accurate to assume that both distances, that for NHK and the BBC, mark the same culturally variable transition point in social distance (I propose to refer to this distance as constructed visual distance, CVD). Without further experimentation, it is impossible to know which might be closer to the truth but both possibilities must be entertained.
There remains the task then of deciding how, and indeed if, CVD is related to typical social distances within a culture. Any comparison between countries has to be contextualized by this knowledge as it is only in relation to the cultural norms that representations can be interpreted as close or distant. If portrayals of social distance (SD) in media are to be compared accurately, then it will be necessary to look at a ratio of ratios; that is, the CVD:SD ratios for two, or more, broadcasters/cultures, where SD is some known type of social distance, for instance the typical distance between conversational dyads. While this information is available for the UK and Japan, inconsistencies in the SD element make firm conclusions difficult.
Further research: Further work on social distance falls within the remit of other disciplines -anthropology, social psychology or certain areas of linguistics. In the meantime, it would certainly be beneficial for media analysts to carry out comparative work looking at variation in CVD across, for example, Figure 8 Scale showing social distance implied by typical NHK and BBC social actor portrayals. Vertical text labels the outer boundaries of Hall's various social distances, the 'public distant' division has no outer boundary and is thus not labelled. genres within one broadcasting system -comparisons of, for example, portrayals in fictionalized and non-fictional accounts of a society. Or between the products of different types of media outlets, for example, public service and commercial broadcasters. Such intra-cultural comparisons, while necessarily circumscribed, would avoid relying on SD data that may or may not exist.
It may also be possible to carry out experiments which use linguistically deculturated news stories as stimuli in an attempt to get at how, for example, Japanese viewers assess the portrayals of social actors presented by the BBC, or UK viewers those of NHK. This would provide a subjective account of perceptions of CVD if nothing else.
This study has chosen to concentrate on one particular element of the televisual image -camera-recorded video. It has therefore left uninvestigated the interrelationships between video and on-screen graphical representations of language such as captions and 'bugs'. The differences in the ways the two broadcasters choose to integrate pictorial and linguistic onscreen elements is certainly worthy of further investigation. Such studies will require a broader theoretical base and necessarily involve consideration of a wider range of influencing factors, at very least, working practices within the production environment and television's relationships with other types of multimodal text.