It’s not enough to be seen: exploring how journalists show aged care in Australia from 2018-2021

ABSTRACT Older Australians, particularly those in aged-care settings, are frequently targets of persistent discrimination and marginalisation. Media portrayals of older people contribute to how broader society sees and values this demographic. Acknowledging this, the present study analyses how journalists visually cover ageing and the aged care sector during a critical event ‘frame’: the calling of, and government response to, the Royal Commission into Aged Care Quality and Safety from 2018 through 2021. This study recognises that the type of representation of older people in media is more difficult to examine than simply the frequency of representation. Using visual social semiotics as an analytical framework, this paper examined 351 images from a nationally representative news sample published over the 30-month timeframe. This approach has enabled us to go beyond simple frequencies of who is depicted and explore in a more nuanced way how older Australians are depicted, and with what implications.


Introduction
About 1 in 6-or roughly 16% of -Australians are 65 or over (Australian Institute of Health and Welfare, 2021). This cohort of older Australians is growing and is projected to increase to between 21% and 23% of the population by 2066 (Australian Bureau of Statistics, 2017). Yet, despite this, older people have historically been marginalised, and are often invisible in media portrayals of issues relating to their health and wellbeing. In this paper, we analyse how journalists visually cover ageing and the aged care sector during a critical event 'frame': the calling, in 2018, of the Royal Commission into Aged Care Quality and Safety, the period covering the Royal Commission's proceedings, and the government's response to the final report in 2021. We begin by tracing how the news media typically cover older people and the topic of aged care in Australia, discuss visual social semiotics as an analytical approach for evaluating the engagement and meaningmaking potential of an image, and detail sampling and data analysis procedures, before presenting the results of that analysis. Such an approach is necessary to better understand how successfully (or not) journalists are using visuals to engage a largely apathetic audience and how the way journalists visually cover ageing and aged care potentially influences how audience members perceive these topics.

Background
One of the first background papers the Royal Commission published (Smith, 2019) noted that a dominant and sustained narrative in Australia was that 'the ageing of the population is seen as a problem to be fixed and that older people are a burden facing the nation' (p. 3). Research shows such perceptions are most commonly held by younger people and are shaped, in part, by media portrayals (Australian Human Rights Commission, 2013). Older people themselves also perceive themselves and their issues as being less visible relative to the rest of the population, which results in 'strong and negative emotional response amongst older people' (Smith, 2019, p. 4).
Neoliberal and capitalist societies value individual responsibility and economic productivity. As such, less value is sometimes placed on older people who are at or past retirement age, when compared to younger people who are still part of the workforce (Rubinstein & de Medeiros, 2015). This conflation of the economic productivity of individuals with their social worth has contributed to the challenging conditions in aged care, including persistent understaffing, skills deficiency, poor and ineffective regulation, and failures of basic care that plague the sector and led in 2018 to the calling of this Royal Commission into Aged Care Quality and Safety (Holland-Batt, 2020).
Media portrayals of ageing and aged care are only one dimension in a complex assemblage of social forces that influence community perceptions regarding older people and their treatment. However, such portrayals are important because not all members of a society have first-hand experience with advanced age or aged care and thus must rely on depictions found in popular culture and journalism to inform their awareness and understanding of these topics. The visual dimensions of such portrayals are especially significant because visuals attract more attention and engagement compared to verbal descriptions (Keib et al., 2018), are more emotionally laden (and therefore more persuasive; Seo, 2020), and are more cognitively efficient (Kumar & Kumar, 2016).
In addition to this, the way that people consume news and information and interact in the digital environment means that the visual dimensions of such news and information messages are potentially influential in shaping views. For example, rather than seeking out news directly from news websites, Australians more frequently rely on social media platforms for news (Park, Fisher, McGuinness, Lee, & McCallum, 2021). In doing so, users view still images in preview links or embedded videos to (1) learn about newsworthy events and (2) to decide whether to click through to an accompanying written account. Images, particularly those that are positively valenced, serve as implicit gatekeepers that govern attention, perception, and engagement (Keib et al., 2018). Due to this, the visual dimensions of news representations, in general, and those related to ageing and aged care, in particular, can play outsized roles in the realm of public opinion and discourse related to these topics.
These gaps aren't necessarily because of structural biases in journalism (such as a predisposition for conflict, novelty, or impact, as the aged care sector has these), but are rather more likely due to social discomfort with the idea of older age and preferences for it remain invisible and out of sight so that younger generations don't have to grapple with its effects for as long as possible. However, it is also worthwhile to note that forces like infotainment can reduce the opportunity for journalists to produce important but complex reporting on topics such as ageing, which are themselves complicated.
Ageist stereotypes in media discourse contribute to an 'us versus them' framing and, potentially, mentality for those who are exposed to such messages (Cruikshank, 2013). Additionally, older people are often physically separated from the rest of society and thus 'othered' (Hazan, 1994;Thayer & Skufca, 2020). This 'othering' can extend into media depictions depending on the various semiotic resources that journalists draw upon when making or selecting photos to accompany the news.
Within Australia, a 2021 study (Imran) of 49 articles related to aged care from 2011 to 13 (published by eight news outlets, including The Australian, The Herald-Sun, The Courier-Mail, The Advertiser, The Daily Telegraph, The Age, The Sydney Morning Herald, and The Canberra Times) found that more than half these outlets discussed the financial implications of aged-care reform more than any other facet of aged care (Imran, 2021). This was followed by political and, in third and last place, social issues. The same study, which primarily focused on written text, noted that 'images are not a central pillar of this project' but considered some key images 'because of their significant contribution to storytelling' (p. 50).
Indeed, much of the extant research on ageing and aged care discourse tends to focus on verbal discourses at the expense of visual ones (see Lichtenstein, 2021), sometimes for pragmatic reasons (eg, that popular news databases for researchers such as ProQuest and Factiva only provide text and not image results), disciplinary conventions, or the historical and ongoing effects of logocentrism (Bock, 2020;Lemke, 2002). The present study responds to calls by scholars (Köttl, Tatzer, Ayalon, & Bowers, 2021) to analyse other types of media (eg, online media), as well as graphics and pictures. It recognises that 'the type of representation of older people in media is more difficult to examine than the simple amount of representation' (Martin, 2009) and, as such, seeks to go beyond simple frequencies of who is depicted to explore in a more nuanced way how these people are depicted and with what implications. Visual social semiotics is one such way that a more nuanced and holistic analysis can be attempted.

Visual social semiotics
Visual social semiotics is a theoretical and analytical framework for 'examining how images convey meaning' (Harrison, 2003, p. 47). Its proponents argue that an image is a social process where meaning is negotiated between the creator and the viewer, and where social, cultural, political beliefs, values, and attitudes can influence the meaning that is made. Indeed, semiotic modes are shaped both by intrinsic qualities of the medium, camera-produced images, in this case, as well as by 'histories and values of societies and their cultures' (Kress & van Leeuwen, 2021, p. 20). Visual social semiotics has its roots in the rhetorical tradition and operates within a system of three image-types: the icon, index, and symbol, though other scholars (such as Rogers, 1989) have proposed broader classifications.
An icon resembles what is represented. For example, a photo of an older person typically looks like the older person depicted but certain viewpoints, camera settings, and equipment attributes can influence the level of individuality (ie, as one specific person or as representative of a more generic type of person). This is the case, for example, with backlit silhouettes where identifiability is potentially lost but salient attributes of the type; for example, hunched shoulders, a bent back, or the presence of a mobility aid, such as a wheelchair or walker, can indicate the image represents a certain type of person rather than a certain individual. Rogers (1989) defines these as 'exemplar' icons (ones that depict a common example of the class of people/place/object they represent) and contrasts these with what she terms 'resemblance icons' (which are direct likenesses of the people, places, or things they represent).
An index doesn't resemble the person, place, or thing being depicted but instead, resembles something that implies the person, place, or thing. For example, a bruise is an index of injury. Depending on the specific context, it can also more narrowly come to represent physical abuse. The correlation between signifier and signified in an index can be known innately or can be learned. A symbol represents an arbitrary association, including things like flags, potentially, and traffic lights. van Leeuwen (1996, 2021) argue that an image performs three metasemiotic tasks to create meaning. These include the representational metafunction, interpersonal metafunction, and compositional metafunction. They describe the interpersonal metafunction as about the actions among all the participants involved in creating and viewing an image. This includes the creator, those depicted, and the viewer. This metafunction answers the overall question, 'How does the picture engage the viewer'?. It operates specifically through four features and processes, including image act and gaze; social distance and intimacy; and horizontal and vertical perspective, which are each unpacked further below. Due to the number of images in the sample (in excess of 350), as well as past research that has found that ordinary Australians are apathetic towards the topic of aged care (Smith, 2019), this study focuses exclusively on the interpersonal metafunction as it relates to how the news depictions produced during the Royal Commission into Aged Care Quality and Safety attempt to engage an audience.
Image act and gaze refers to the gaze of the person(s) depicted in relation to the camera and thus, the viewer. A 'demand' image act and gaze occurs when the person(s) depicted look directly at the camera and, thus, the viewer. Conversely, an 'offer' image act and gaze occurs when the person(s) depicted look outside the frame or at someone or something within the image. The person(s) are shown more passively with this image act and are presented as an object of contemplation, leading to less engagement than with a 'demand' image act.
Social distance and intimacy refers to how close the person(s) depicted are to the viewer, which connotes feelings of intimacy or distance. Harrison (2003) defines six potential social distances, including intimate (where the head and face only are seen), close personal (where the head and shoulders are seen), far personal (where the figure[s] are seen from the waist up), close social (where the whole figure is seen), far social (where the whole figure with space around it is seen), and public (where torsos of several people are seen). This study added an additional, seventh category of other/mix for images that showed people at multiple planes or grounds (foreground, midground, background) within the same image.
Horizontal perspective refers to the positioning of the subject(s) in the frame relative to the camera and the viewer. When those who are observed are shown frontally to the viewer, this angle creates stronger engagement with the viewer and implies that the person(s) depicted are 'one of us'. Conversely, when an oblique angle is used, this creates a sense of detachment and implies that the person(s) depicted are 'one of them'.
Vertical perspective refers to either the vertical relationship between the person(s) depicted and the camera/viewer or between or among the people depicted in the frame. For reasons of scoping the data analysis, this study adopts the former operationalisation and examines only the vertical angle between those depicted in the frame and the camera and viewer. Therefore, the consideration is of the meaning of angles, such as high camera angles where the person(s) are 'looking up' which tend to show the person(s) depicted with less power. Medium-angle perspectives where the person(s) depicted are looking 'horizontally' tend to show them with equal power to the viewer. Low-angle perspectives, where the person(s) depicted are looking down, tend to show them with greater power.
The study's umbrella research question is interested in how journalists deploy various semiotic resources in their visual coverage of aged care during the Royal Commission into Aged Care Quality and Safety to engage audiences and potentially shape their perceptions of ageing and aged care. It asked: Across a nationally representative sample of journalists' coverage of aged-care related news during the Royal Commission, how did journalists deploy each of four semiotic resources (image act and gaze; social distance and intimacy; and horizontal and vertical perspective) within the interpersonal metafunction?
This article uses 'journalists' as an umbrella category for all news workers involved in the media production, editing, and publication processes. As such, it includes photojournalists and reporters with smartphones who are creating visual media and photo or social media editors who are often responsible for selecting which audiovisual footage and images to use, edit, and publish (Schwalbe, Silcock, & Candello, 2015). It acknowledges that the process of gatekeeping pervades each stage of the process: from embodied gatekeeping in the field that determines who is allowed access a particular scene (Bock, Istek, Pain, & Araiza, 2018) to gatekeeping forces by editors, as acknowledged above, as well as gatekeeping forces by algorithms that influence what is seen, when, and by whom (Wallace, 2018). As such, the study recognises that journalists are not always afforded complete autonomy in their decision-making regarding production, editing, or publication factors but that this potential lack of autonomy does not detract from the importance of paying close attention to the visual news representations of older Australians and how they are being covered.
The study's four sub-research questions are more granularly interested in how journalists deployed each of these four semiotic resources in relation to the types of people who were depicted. They ask:

Data collection procedures
In order to determine how best to analyse how journalists show the story of aged care in Australia during the Royal Commission into Aged Care Quality and Safety, the research team first considered the broad history of policy developments that have shaped the aged care sector in concert with the (decreasing) quality of care provided that led to the calling of the Royal Commission into Aged Care Quality and Safety in October 2018. This milestone was selected as the beginning of the study's timeline. The Commission's final report was tabled on 1 March 2021, and the government provided its response to it on 11 May. The study's end timeline was selected as 11 June 2021, approximately four weeks after the government's response, in order to capture any additional news coverage that followed in the wake of the government's response.

Sampling considerations
The study employed a purposive sampling strategy (Etikan, Musa, & Alkassim, 2016) that included both national and regional news coverage on the topic of aged care in Australia during this timeline (1 October 2018, through 11 June 2021). Specifically, it selected 13 outlets overall, including three national news outlets (The ABC, The Australian, and The Guardian Australia) and 10 regional news outlets (the Sydney Morning Herald, the Daily Telegraph, the Canberra Times, The Age, the West Australian, the Brisbane Times, the Courier Mail, The Advertiser, The Mercury, and The Northern Territory News). In this way, the study achieved representation from each Australian state/territory, and a diversity of media ownership models.
Articles were selected based on a keyword search that consisted of the following three sets of keywords: 'Royal Commission' and 'aged care', 'home care', and 'elder abuse'. 1 The lead author consulted a nationally leading Australian aged care advocate to determine these keywords and to ensure that the coverage would be inclusive of both residential care (which accounts for about two-thirds of the sector) and home care settings and would include all coverage related to the Royal Commission, the aged care sector, and related issues.
The databases used to compile these articles (Factiva and ProQuest) only provided text results and not any accompanying images so the lead author, over a period of 10 weeks (from the last week of June through the first week of September 2021), individually looked up each article and downloaded the lead image (if any was present) that appeared in it. The rationale for this is the lead image is almost always the one that is shown when article links are shared on social media. Therefore, these are the images most likely to be seen and that represent the story of aged care in Australia -even if viewers didn't click through to read the linked article. Because only single images were collected and analysed (rather than photo galleries) and because mugshots were not typically used as lead images, the images in this sample most closely align to the 'news photography' genre Kędra (2016) proposed.
There were a total of 381 images with identifiable people in the sample. Of these, 30 were stock photos and removed, as the intent of the study was to examine the story of aged care during the Royal Commission in Australia rather than examining depictions in stock photography banks that were likely made in other contexts using other conventions. 2 We argue that stock photography functions as 'conceptual images' (or what Kędra, 2016 calls 'illustrative photography'), which Kress and van Leeuwen (2021) define as stylised images with 'conventional symbols to illustrate the essence of an issue, rather than documenting newsworthy events' and contrast these with '"uncoded" naturalistic representations of events', such as those produced by journalists or family members. (p. 29). Based on this definition, stock photographs were determined to be qualitatively distinct enough to warrant their exclusion from the analysis. As such, once these stock photographs were removed, the final sample included 351 images. Of these, 18 (5.1% of the overall sample) were duplicates that were published multiple times (either by different news outlets owned by the same conglomerate or by different outlets with access to the same visuals). Forty-nine images (13.9% of the sample) originated from wire services. These included 43 from the Australian Associated Press (AAP), one from the Associated Press (AP), and five from NewsCorp's NCA Newswire. Thirty-one images (8.8%) were 'supplied' and were sourced from family members, from Royal Commission materials, or from other sources. And 10 images (2.8%) were screen captures from a Royal Commission livestream or from a television broadcast.

Operationalisations regarding the horizontal perspective category
In a conventional news image with an identifiable subject, the angles available typically range from 0 degrees (when the subject is parallel to the camera) to 90 degrees (when the subject is shown in full profile). Technically, the frontal angle is 0 degrees and the oblique angle is, then, anything from 1 to 90 degrees; however, due to this wide variance and the desire to ensure that the depiction provided a clear sense of engagement or detachment, this study operationalised the oblique angle to be anything from a three-quarters to a profile view (approximately 45-90 degrees) and operationalised the frontal angle as anything from 0 to 44 degrees. This exact threshold is admittedly a subjective decision and the results could have differed had a less distinctive and more subtle (but arguably less impactful) angle been chosen.

Analysis
Analysis began with a deductive coding process that first explored the prevalence of semiotic resources across the overall sample. These are reported in the first part of the findings section that follows. Following this, these same four semiotic resources were examined in relation to people types (a type of axial coding) and all possible people types and combinations were identified and coded (eg, older people shown with aged care workers, older people with family members, etc.) however, images that showed different people types within the same frame (e.g., older people alongside aged care workers) muddled the meaning that could be inferred from the analysis so only image groups that depicted one type of person in the frame are reported in the analysis that follows. Additionally, because of the large number of categories (n = 20), including many with low frequencies within them, only the top five people types are reported in the subsequent analysis. These five people types accounted for nearly three-quarters (73%) of the entire sample.

Findings
The study's umbrella research question was interested in four broad semiotic resources (image act and gaze; social distance and intimacy; and horizontal and vertical perspective) journalists employed to engage audiences in their visual coverage of aged carerelated news during the Royal Commission into Aged Care Quality and Safety. It asked: Across a nationally representative sample of journalists' coverage of aged-care related news during the Royal Commission, how did journalists deploy each of these four semiotic resources to engage an audience? Overall findings will be shared for each of these resources first and then will be followed by the sub-RQs that addressed how the resources differed across different types of people.

Image act and gaze (overall)
Related to the first semiotic resource, image act and gaze, the journalists in the sample relied most heavily on the 'offer' image act (in 56.1% of sample) where a lack of eye contact existed between the person being depicted and the camera/viewer (see Figure 1). 'Demand' image acts, where the viewer makes direct eye contact with the person(s) depicted, took place in slightly more than 40% of all the images. 'Other' or 'mixed' image acts (where some people in the frame stare directly at the camera while others do not) occupied the minority of the sample with only 12 instances (3.4% of the overall coverage).
It is necessary to not just interpret meaning superficially at the content stage but also to consider potential influences at the production and editing stages, including environmental factors, interpersonal dynamics, and the effects of routines and organisational forces, on shaping coverage and ultimately influencing meaning (Thomson, 2019). As such, it is relevant to note that the proportion of 'demand' photos in the sample, overall, seems high considering industry conventions such as those discussed by Kobré (2017), who writes: Photojournalists must catch their subjects as unaware as possible to record real emotions . . . In good candid pictures, subjects never gaze at the camera. Eye contact tips off the reader that the picture is not candid and suggests that the subject was at least aware of the photographer and might even be performing for the lens (p. 28).
Different conventions exist between news imagery which generally strives for candid and organic moments and other types of photography, such as personal and family snapshots where posing is expected and encouraged. Such conventions are relevant from an engagement standpoint considering past research, using eye-tracking methods, (Leckner, 2012;Quinn, Stark, & Edmonds, 2007), that found that 'documentary photos receive more visual attention than staged photos' (p. 173). People are more familiar with posed personal and family snapshots, which likely affects how they are perceived compared to the more unique candid depictions that are a hallmark of photojournalism.

Social distance and intimacy (overall)
Of the seven different social distance and intimacy categories that this study uses, two of these -head and shoulders (close personal distance) with 38.5% of the sample and from the waist up (far personal distance) with 33% of the sample -dominated the news coverage over these 30 months and accounted for more than two-thirds of all the depictions (see Figure 2). No other category comprised more than 10%. In descending order of prevalence was showing the whole figure with space around it (far social distance) with 8% of the sample; showing the head and face only (intimate distance) with 7.1% of the sample; showing torsos of several people (public distance); and each with 4.3%, showing the whole figure (close social distance) and showing a mix of people within the same frame at various distances and intimacies.

Horizontal perspective (overall)
With the more generous operationalisation of 0-44 degrees, images that were primarily frontal to indicate to the viewer that those depicted were 'one of us', occupied the lion's share of the sample with 275 instances (78.3%) (see Figure 3). Images that were firmly in the oblique realm (from 45 to 90 degrees, a three-quarters perspective to a full profile view) followed with 58 instances (16.5%). A mix of perspectives due to multiple people in the frame at various angles rounded out this category with 18 instances (5.1% of the coverage).

Vertical perspective (overall)
The journalists in the sample were on an equal vertical plane with those they covered in the majority (58.4%) of cases (see Figure 4). They put those they were photographing in a privileged position (through a low-angle perspective) in 18.8% of cases and put those they were photographing in a subservient position (through a high-angle perspective) in 22.8% of cases. Again, journalists do not necessarily enjoy full autonomy related to the angles, vantage points, and perspectives they can adopt. For example, journalists photographing in Parliament House in Canberra are, according to the government's guidelines (last revised in 2008) barred from using telephoto lenses to show any Member or Senator  'any closer than at "head and shoulders" distance' (Parliament of Australia, 2008, p. 13). Similarly, the vantage point journalists can work from is sometimes limited by event organisers and this can, in turn, affect the angles possible.

Image act and gaze (by type of person)
The visual social semiotic analysis was also cross-referenced with the types of people depicted.
The top five most prevalent people types shown, and how they were represented is shown in Table 1. The most commonly depicted were politicians, with 28.2% of the sample, followed by older people (21.4%), public servants (8.8%), businesspeople (8%), and older people's family members occupying the remaining 6.5%.  Importantly, the distribution of image act and gaze by type of person was not equal across the sample, varying significantly for older people and their families versus the other groups. Journalists favoured the 'offer' image in a majority of cases, showing politicians, public servants, businesspeople, and older people's family members most often looking beyond the camera. Only with the older people and older people's family categories did journalists show them more frequently with the 'demand' image act.
The high proportion of 'offer' framing in the images of politicians is perhaps not surprising since one of the key functions of journalism is to serve as a watchdog to power (Knobel, 2018). News depictions that seek to candidly document politicians and their actions (rather than allow them to pose for the camera and control elements of their framing) is evidence of this function.
The high proportion of all the people being depicted as more agentic through the 'demand' framing might additionally appear a positive finding. However, almost half of the depictions of elderly people sourced (using the above described sampling strategy and across media) were originally from family members or from the Royal Commission itself and most often vintage family photos. That is, these 'supplied' photos were 31 of the 75 depictions, 41.3%. Within these supplied pictures the 'demand' image act was featured in 23 (74.2%), 'offer' in 7 (22.6%) and a 'mix' was featured in a single image (3.2%). Nevertheless, even if these supplied photos were removed from the sample, the number of 'demand' images featuring older people would still occupy more than half (56.8%) of the images featuring this people type (dropping from 48 to 25). Journalists afforded older people's younger family members with the most agency and featured them with the 'demand' image act in nearly three-quarters (73.9%) of cases.

Social distance and intimacy (by type of person)
Overall, journalists in this sample preferred tighter framing as the 'head and shoulders' social distance and intimacy category was the most prevalent across the sample and was the dominant one for the politicians, older people, public servants, and business people in the sample. Only when depicting older people's family members did journalists depict them from the slightly more distant 'waist up' framing (often because there was more than one person to depict or because, in the case of solo figures in the frame, they often posed alongside framed photos or signs and including these in the shot required wider focal lengths). The extremely tight and intimate 'head and face' category was the fourth-most popular and was deployed most often when depicting older people.

Horizontal perspective (by type of person)
The data show that frontal perspective was deployed in an overwhelming majority of all the five most prevalent people categories. Most significantly, older people's family members were shown as 'one of us' through frontal framing in 95.6% of cases.
All of the five most prevalent people categories were also shown with an oblique 'one of them' angle; however, the distribution amongst these top five was not even. Business people, a category dominated by people (such as Aveo executive Andrew Macintosh, Estia Health chairman Gary Weiss, and Medibank CEO Craig Drummond) were the most likely to be visually shown as 'one of them' with the oblique angle occupying 25% of all depictions in this category. This was followed by depictions of public servants (such as Commissioners Lynelle Briggs, Janet Anderson, and Richard Tracey, and Department of Health Secretary Prof Brendan Murphy) from an oblique angle, with 19.3%. Both the elderly and politicians were visually shown as 'one of them' in 14.6 and 13.1% of cases, respectively. The least likely to be visually shown as 'one of them' was the (nearly always younger) family members of older people, who were only shown obliquely in 4.3% of cases. Interestingly, the journalists choose to depict business people (most often CEOs or executives) as 'one of them' perhaps not only because of their higher socioeconomic status compared to the earnings of an average Australian but also because they themselves are also older Australians (Macintosh, Weiss, and Drummond, for example, were all in their late 50s or 60s when their depictions were made.)

Vertical perspective (by type of person)
Journalists in the sample favoured the 'medium' vertical perspective evenly across the board regardless of the type of person depicted. This was unlike the other two semiotic resources of image act and gaze and social distance and intimacy. It places the person depicted eye-to-eye with the camera (and thus the viewer), making it an egalitarian and equal framing that is the most neutral of the vertical perspective possibilities.
Importantly, beyond this most prevalent medium perspective, journalists relied almost three times as much on the high vertical perspective when photographing older people. As explained above, this framing places those depicted in a less powerful, subservient position. For framing politicians low or high perspectives were equally likely to have been used. Public servants and older people's family members tended to be shown in more powerful ways -from lower angles. For business people there appears to be a slight preference for framing them in less powerful ways (from high perspectives).
Although word limits prevent a full exploration here, we argue that the engagement potential of an image is more potent when multiple semiotic resources are compounded. For example, when a 'demand' image act is paired with a low vertical angle so that the depicted is shown as a pleading demand. Such compounded effects are also the case when a 'demand' image act is paired with a high vertical angle and is shown to tell the audience what to do.

Discussion
Journalists in this sample deployed the semiotic resources of image act and gaze; social distance and intimacy; and horizontal and vertical perspective in varying ways. Encouragingly, journalists didn't uniformly offer up older people for passive inspection but showed them most frequently as active and agentic. However, journalists also showed older people from a greater social distance more often than any other people type. Older people were also the ones journalists showed most frequently with the least amount of power through the vertical camera angles that were used. This means that in some ways journalists are still visually 'othering' older people by showing them at greater social distances and as being 'one of them' compared to younger people. It also means that enduring stereotypes, such as older people being dependent and weak, are channelled into the camera angles that journalists use.
Journalists concerned about viewer engagement and attention relative to a marginalised topic like aged care should consider whether they are using semiotic resources arbitrarily, for ideological reasons (e.g., to comment on power relationships), or pragmatic ones (e.g., to clear up a distracting background or to use oblique lines to enliven an otherwise static composition). Journalists seeking to cover the plight of older Australians in aged care should also consider the effect of portraying individuals over groups in the light of psychological principles, such as psychic numbing (Slovic, 2010) and innumeracy, where humans' ability to care or appreciate large numbers is secondary to their ability to appreciate and comprehend the plight of a single individual. This depth, not breadth, approach is more likely to generate an empathetic response and to relay the possibility of hope over the seemingly hopeless reality a more macro and distant perspective might offer.
Relevant, too, is a discussion of representativeness. The Commission's interim report, titled 'Neglect', found that Australia's aged care system 'did not deliver uniformly safe and quality care, is unkind and uncaring towards older people and, in too many instances, it neglects them'. If the majority of those in aged care are neglected, isolated, receiving insufficient care, and are lonely, journalists must ask themselves whether it is their role to reflect this reality as it is currently or if they should seek out more caring and humane anomalies in order to portray a wider future that might be and, through that speculative future, seek to help shape perceptions about ageing and aged care for current and future generations.
Worrying about a particular threshold or ratio of one semiotic resource to another within one's imagery is not enough. Context is key and deserves to be considered. Not only is the type of person who is being shown relevant to consider but also the circumstances of that presentation. If the image's purpose is merely identification and to show what the person looks like who is quoted (as in the case of an often vertical or square headshot that is embedded within the article as a secondary element), this image serves a different purpose (and has a different engagement potential) than an often horizontal 'hero' one that is the first to be seen, is often the largest, and is the one that is shown when the accompanying article is posted to social media platforms. In summary, a 'demand' image isn't always more engaging than an 'offer' one, depending on the context.

Limitations and future directions
While this research is novel, it is also generally exploratory. Regarding its sample, it does not analyse all types of news depictions related to ageing and aged care but only those that are present in 13 news outlets in a single country. Different kinds of storytelling (television compared to photojournalism, for example) have different conventions and the nuances of these conventions and genres deserve to be studied as they relate to under-represented attributes such as older age. Likewise, this study's sampling frame takes place during a Royal Commission, which is a period of heightened attention to and awareness of aged care by some stakeholders. Such events can artificially influence which stakeholders are shown in the news and in what ways compared to other periods of coverage where a government spotlight isn't shining on the topic.
More granular analysis would also be welcome on specific identity attributes within aged care coverage, such as how these semiotic resources compare across gender, ethnic background, and class. A further opportunity exists to study multimodal elements in concert with one another, such as how the meaning of the photo compares to the meaning offered in the caption and/or headline of the accompanying article.

Conclusion
This study responded to gaps in the literature and calls for more research that addresses the visual side of ageing and aged care-related issues and, by doing so, offers a concrete sense of whether existing stereotypes and trends that scholars have identified in news and media depictions, broadly, also extend into the visual realm and how these depictions can be nuanced to engage an apathetic audience. Various identity attributes, such as race, gender, or religion, can be subsumed by the category of age once a certain age threshold is met yet older people also find themselves in the margins of society and deserve additional attention and consideration. A society's worth can be determined, in part, by the value it places (or fails to place) on its most vulnerable members and this extends to how these people are shown in news and media.
Ultimately, journalists have a unique responsibility to engage audiences about older people in aged-care settings. Visuals present an excellent opportunity to do so through careful consideration of their meaning-making interpersonal metafunction. However, not all engagement is equal and such engagement doesn't necessarily translate to empathy. Faces engage us (Bakhshi, Shamma, & Gilbert, 2014;Torbarina, Jelenc, Brkljačić, & Jelenc, 2019) but engagement for engagement's sake should be avoided if it is voyeuristic or simply reproduces existing stereotypes and tropes. Rather, by considering not only who is represented but how they are represented, journalists can harness the power of visual media to raise awareness, cultivate engagement, and help shape perceptions related to ageing and aged care.

Notes
1. Not all keywords returned results. For example, the keyword 'elder abuse' did not return any results but all keywords used are still reported for full transparency. 2. Even though conventions like whether the action is posed differs markedly between stock photography and most types of news photography, it is possible that awareness of visual tropes and stereotypes prevalent in stock photography libraries, as well as in other popular culture sources, such as movies and television shows, also affects how journalists see older people and choose to represent them.

Disclosure statement
No potential conflict of interest was reported by the author(s).