Introduction

In 2014, a leaked internal report from The New York Times expressed dismay with editorial fixation on the front page and over investment in its print legacies: ‘The newsroom is unanimous: we are focusing too much time and energy on Page One.’¹ Yet, despite this, as the report notes, it ‘persists’: ‘Page One sets the daily rhythms, consumes our focus, and provides the newsroom’s defining metric for success.’² The front page remains a site of tension between the paper’s ‘digital needs,’ portrayed as crucial to its long-term sustainability, and its ‘print traditions,’ seen as a negative, conservative force.³ Parallel to this focus on the digital, a prominent tension between image and text at the newspaper has emerged in its corporate literature, with calls for a ‘more visual daily report’ instead of ‘long strings of text’ and with emphasis on providing photographers with ‘the primary role’ in creating stories.⁴ As such, the front page provides a symbolic battleground with which to understand larger trends within the industry: tradition versus innovation, print versus digital, and text versus image.

These tensions have been intensified through a post-2000s period of dramatic financial destabilisation for both newspapers and photojournalists. Catalysed by the rise of internet distribution in the early 2000s, the precipitous fallout of advertising following the 2008 Financial Crisis decimated newspapers internationally, which were already struggling with declining circulation numbers from their heights in 1990.⁵ Given the scale of its impact worldwide, this rupture was swiftly followed by a flurry of books and articles declaring the death knell of newspapers.⁶ The Times’ outlook has been consistently gloomy. Marie Bénilde’s 2010 op-ed ‘The End of Newspapers?’ contended that ‘Journalists are now in the same situation as steel workers in the 1970s: They are destined to disappear, but they don’t know it.’⁷ In August 2020, outgoing New York Times Company CEO Mark Thomas predicted the paper will stop their print edition within the next twenty years, focusing on its journalism-related intellectual property (recipes, romance columns, and school curriculum) and publishing solely digital content.⁸ Yet, the newspaper has oddly benefitted from the disruption. According to Robert W. McChesney, the paper can credit its outsized influence as a cosmopolitan newspaper in part to the industry’s shrinkage:

In the Internet era, the New York Times probably plays a much larger role in national and international journalism than it ever did in previous generations, despite its own significant cutbacks. It does so because most of the other major news media have abandoned their networks of national and international bureaus altogether.⁹

This sentiment is echoed on a domestic level by Daniel R. Schwartz in his assessment of the newspaper’s resources for international bureaus and correspondents compared to other American news media.¹⁰ As local newspapers folded, The New York Times expanded its newsroom to a record 1,700 employees, well over its pre-2008 staff.¹¹ While the number of bureaus and on-staff foreign correspondents for many newspapers have decreased in this period, The New York Times has taken pains to brand itself as cosmopolitan (as per the word’s literal etymology), as ‘true citizens of the world’ and ‘global thinkers.’¹²

For the New York Times, the last two decades have also represented a time of unprecedented visual creativity. Leading the photography department from 2004 to 2018, Michele McNally has been particularly credited by the paper for spearheading a new chapter for its visual content, pushing ‘a reluctant newsroom’ from utilitarian, minimal editorial illustrations towards photography as a crucial and compelling storytelling element in the newspaper, reaping more awards ‘than most news organizations have won for their entire reports.’¹³ McNally’s impact on The New York Times coincided with wider aesthetic trends in photojournalism gaining mainstream editorial traction. The term ‘fine art photojournalism’ was coined by Julian Stallabrass to express the increased presence of photojournalists in art galleries due to the decline in traditional modes of distributing photojournalism; as such, this mode of photography tends to reject ‘straight’ documentary for highly mannerist styles.¹⁴ This trend for ‘artistic’ styles in photojournalism has also been explored in a general sense by Jennifer Good, Paul Lowe, and Fred Ritchin, and through more focused articles by Marco Solaroli looking at newspapers.¹⁵ Building upon work like this, my own doctoral research examines photojournalistic work from 2000 to 2020 as a heightened period for photojournalism’s broad, visual shift towards increasing subjectivity and what I term as aesthetic hybridisation: the adoption of commercial, conceptual and formalist visual strategies by photojournalists to appeal towards diverse platforms, purposes and audiences.

The front page of The New York Times has transformed since 2000, despite the bleak assessments from inside and outside the paper that its focus on tradition has impeded innovation.

Predictably, there have been many books and articles about this particular paper—including larger academic studies looking at digital photojournalistic content.¹⁶ These, however, have tended to focus on specific case studies (a single issue or article)¹⁷¹⁷ or examine iconic images that continue to resonate with audiences today.¹⁸ David Shields’ 2015 book took a different approach tailored for a general audience, putting together poetic, themed groups from the Times’ front pages between 1991 and mid-2014, but it focused specifically on the paper’s representation of armed conflict.¹⁹ My study also examines the Times’ ‘Page One,’ but unlike previous studies, I am exploring ‘everyday’ photography rather than seeking out any particular genre, style or subject. Through the qualitative coding of 240 digitally archived front pages from January 2000 to January 2020, I move beyond the content of these images to consider trends in formal qualities and institutional context. This article employs visual analysis using close readings of formal qualities combined with a sort of ‘medium viewing’ (which I define below) to gain a sense of the general aesthetic shifts of the paper.

The ‘everyday,’ it must be mentioned, is a term loaded with ideological connotations. In his Critique of Everyday Life, Henri Lefebvre views everyday or lived experiences as crucial to understanding society—optimistically, he suggests that engaging with the everyday through culture can spur the desire for radical societal transformation. Equally, he argues that the de-contextualisation of the everyday—the ‘art of presenting the everyday by taking it from its context, emphasizing it, making it appear unusual or picturesque and overloading it with meaning’—risks its integration in superficial narratives.²⁰ As John Roberts points out, Lefebvre does not offer a prescriptive analysis, but emphasises how the everyday is primarily a site of contradictions, from which ‘artistic and theoretical activity is actually performed out.’²¹ Here, the ‘everyday’ is used to refer to the scope of this study, the daily newspaper. By sampling the first front page of each month, I have consciously tried to avoid curation or a tendency towards the iconic. It is also, however, the ‘everyday,’ and all the ephemerality associated with it, that is in tension with the permanence of the archive, providing many of the inconsistencies I discovered. My analysis generated several questions: What aesthetic shifts, if any, can be found through an examination of everyday news photography, rather than the iconic, from the early 21^st century? How has the status of the image changed in The New York Times in this period? And, how are the tensions between image/text, digital/print, innovation/tradition and everyday/permanent reflected in the Times’ various digital archives?

Methodology

Amid restrictions to in-person resources during the Covid-19 pandemic, the methodology of this piece was necessarily tailored to available online archival sources; in doing so, several conflicts emerged between the physical and digital archival processes. There are a handful of official digital resources, each with their own limitations. The New York Times has two main sources for viewing past newspapers as issues rather than single, online-only formatted articles: the TimesMachine, which is available to subscribers, and ‘Today’s Paper,’ which shows articles and images of the front pages of New York, national, and international editions, which is freely available. TimesMachine covers 1851 to 2002 and presents the papers as grainy black-and-white scans—a tell-tale sign of scanned microfilm—of unspecified but apparently New York, Late Edition newspapers with OCR (optical character recognition) capability to search and browse the texts. ‘Today’s Paper,’ while not a formal archive, nevertheless provides high resolution colour PDFs of the front page in all three editions from around 2012 onwards, and irregular and small, low resolution colour JPEGs for the rest of the 2000s.²² Another digital archive available to researchers showing content in the form of the original, printed newspaper is ProQuest Historical Newspapers: The New York Times (1851-2016). However, no specific, consistent overview is given regarding the editions scanned. ProQuest also supplies grainy black-and-white images with OCR capabilities; photographs are occasionally excised from the paper for copyright reasons and are rendered through the digitisation process (digitised from microfilm, not the paper itself) in high black and white contrast, making many of the images illegible. In a 2003 review of the reference service, ProQuest’s ‘digitization of the microfilm’ is touted as producing ‘better quality copy than the microfilm original.’²³ A scan from microfilm cannot, naturally, provide any more image details from its original, though the process may apply higher contrast values in order to make the texts more readable. The historical hierarchy of text over images in journalism and the subsequent value of photojournalism and its authorship is reflected in these artefacts left over from older archival methods, digitised in the early 2000s.

Data of this scale could arguably justify an automated approach. Improved algorithmic abilities to scan photographs for usable and accurate data have given researchers the ability to identify broader trends on an inhuman scale. This machine ability to ‘see’ images has entered into both academic studies within the humanities and into popular culture. Arnold and Tilton developed the framework of ‘distant viewing’ precisely for this purpose: breaking down the visual language of the image.²⁴ They propose identifying semantic aspects such as dominant colours or recognisable objects and then studying the patterns of these coded elements. Wevers’ and Smits’ work further describes the use of convolutional neural networks (CNNs) to process visual information, applying them towards nearly a million newspaper images and advertisements from 1860 to 1995.²⁵ Both groups of researchers advocate collaboration between technical and non-technical visual experts, but Manovich’s and Tifentale’s research on what they term ‘competitive photography’ on social media networks merits particular attention for recentring the humanities within digital humanities. As they write, ‘Computer scientists treat social media photos as easily accessible data that can be analysed algorithmically.’²⁶ Their study is still enormous, with over 14 million images from Instagram, analysed in part by computational methods. But they highlight the further layer of interpretation they can offer from the humanities, with a grounding in broader or more historic cultural trends that provides context to algorithmic results.

A significant barrier to applying an algorithmic approach, in the case of the identified Times archives, is quite literally the assumption of ‘easily accessible data’: both the informal and formal reservoirs of material I have identified do not have sufficient image quality, with patchy, inconsistent, and sometimes non-existent displays of images. As a researcher, I was forced to triangulate between the three sources to get the basic information I required from both image and text. Even if the archive was more appropriate for a ‘distant viewing’ approach, many of the existing CNNs are still fairly limited and commercially oriented in their aim: an algorithm being able to accurately identify a dog, for instance, is extremely useful for large stock image libraries but perhaps less useful for more complex categories of style and more subtle tropes, which require a deep understanding of images’ cultural and historical context. Even in basic object recognition, CNNs are still prone to errors; over time, no doubt, these computational mistakes will become rarer and their ability to accurately recognise complex patterns will grow.²⁷ Other macroscopic approaches, such as analysing metadata or captions through OCR, equally pose problems to image-based studies. The Times itself has encountered a similar lack of precision in identifying objects and people in their own photography CMS (content management system) as recently as 2020, which runs searches based on caption information and tags.²⁸ Equally, a typical art historical, formalist approach to visual analysis with close readings of individual materials also has significant pitfalls. While close reading works well for more focused studies of iconic images, it is less useful for making sense of broader trends in everyday images.

In this article, I pair manual qualitative coding of a systematic sample (the first day of each month) with close readings of individual images to result in a sort of ‘medium viewing.’²⁹ This lens facilitates looking at meaningful patterns in style in consideration of accumulated cultural history. Though unable to trawl through the same vast quantities of images as Wevers and Smits, this type of ‘medium viewing’ offers an alternative for substantial and complex datasets. With a more consistent and accessible dataset, this approach could effectively complement large-data overviews, as with the cultural analytics in Manovich and Tifentale’s research. This method can draw attention to details that may be overlooked due to the scale of macroanalysis, the lack of consistent or lossless image files, and the relative simplicity of existing CNNs. From a practical point of view, several of the aesthetic developments I outline below are not currently possible to identify through algorithmic analysis. My study of aesthetic qualities in photojournalism contributes a relevant and lesser explored perspective in a research field saturated with industry figures and scholars from digital humanities, journalism, media studies, and sociology.

This article examines the front pages of the newspaper’s national edition, covering the first day of each month from January 2000 to January 2020—240 front pages in total. Drawing from all three digital sources, I analysed the images on the top half of the front page, the part visible when folded and presented on a newsstand, using qualitative, manual coding of 240 front pages from January 2000 to January 2020. When analysing each front page, I recorded information from the by-line, caption, subject and article context, as well as how the picture relates to the layout of the entire page, looking at both black-and-white and colour sources for each cover. If there were multiple lead images, I recorded information regarding all of them. I also recorded if there were no lead photographs (e.g., when an infographic was used). Additionally, I noted the visual character of the image—and whether it fell within the familiar conventions of traditional editorial photojournalism or whether it demonstrated other aesthetic affiliations.

To provide a source of comparison, I sought a popular handbook that sets out some of the normative conventions of 20^th century photojournalism. Many formal texts could be selected for this purpose—from Arthur Rothstein’s 1956 Photojournalism to Jennifer Good and Paul Lowe’s 2017 Understanding Photojournalism.³⁰ As I wanted to consider changes post-2000, I chose a text from the latter half of the 20^th century, Harold Evans’ Pictures on a Page, first published in 1978, for its particular relevance to editorial photojournalism. Evans served as editor for the British newspaper The Sunday Times and later went on to leading international newspapers and magazines. In his handbook, Evans reviews practical techniques and design strategies for capturing a newsworthy image. In the 1997 revised edition, Evans remarks in a new preface that although ‘photojournalism has been given wings in the twenty years since the first edition of this book’ in terms of the speed of the internet and the capabilities of computer editing, the ‘purpose of Pictures on a Page remains the same. It is for the people who put pictures on pages; and for all of us who look at them in newspapers and magazines. I hope it will stimulate the professional and entertain the ordinary reader.’³¹ I selected Pictures on a Page precisely for the conventional and institutionalised perspective it offers on photojournalism, providing a broad understanding of the status quo with which to measure new trends, rather than, for instance, the critical studies on the ‘grand narratives’ of photojournalism and documentary photography that arose around the same time.³²

Findings

Though my sample of 240 front pages cannot reveal any significant statistics on the scale of macroanalysis, some numbers illustrating the constitution of this dataset provide useful institutional context. In terms of the types of photographers, the patterns I find in my study of The New York Times mirror the aforementioned accounts from journalists regarding the destabilisation of the field. Many of the images pre-2008 suggest a relatively healthy industry for wire and staff photojournalists, whereas the post-2008 reveals the prevalence of freelancers. Seventy of the front pages had images taken by staff photographers—a total of 23 different staff photographers shot the photographs. Most of these images appeared between 2000 to 2010, with a decline in photographs by staff photographers after this period. Additionally, of the staff photographers who do appear post-2010, only four new names appear, demonstrating a lack of new hires. In total, photographers from major wire agencies such Associated Press, Agence-France Presse, and Reuters provided 74 of the front-page images, while freelance photographers not associated with an agency featured on 63 front pages. Another 27 front pages did not name the author of a photograph. Most of these were credited only to wire agencies, taken between 2000 to 2004, and related to an international (non-US) crisis, such as a natural disaster or conflict—a situation where a bureau photographer would have less access to a breaking story. The rise of named, individual authorship suggests a shift to the photographer as a true author rather than simply ancillary to the more important article by-line.

The images also indicate an increasing disappearance of local stories (local politics and interest stories) from the front page between 2000 and 2020, equally mirroring The New York Times’ stated objective of appealing to an international readership. After the early 2000s the paper appears to favour more nationally inclusive domestic stories around federal politics and international and domestic crises such as natural disasters and conflict. The lack of difference between the ‘local’ and ‘national’ editions also attest to this. Over half of the front pages featured works by American or dual-national photographers (146/240), and there was an equal split between international and domestic subjects being depicted (111 to 118). There were several stories related to US military presence in Afghanistan and Iraq over this period, but 50 countries in total were represented on the front page. In this respect, The New York Times’ positioning as a global brand becomes even more evident. Published in January 2017, a group of journalists for The New York Times released a report, detailing their longstanding goals and aspirations for 2020.³³ Crucially, the team advised a new balance of image to text as a goal to explicitly build their audience globally.³⁴ This reflects a pronounced ‘cosmopolitan’ characteristic of news in the 21^st century that has been widely theorised by scholars.³⁵

Visually, the most striking difference between the 240 front pages from 2000 to 2020 is the space given to images on the page in relation to the text. Around 2000, the lead images are displayed as small as 1/8 or 1/10 of the page, often providing a portrait insert or conventional illustration to a story either on the same page or further inside the paper. Around the late 2000s, stand-alone images appear, unconnected to any story and with a short caption explaining its significance. From 2006 onwards, image size increased to around 1/6 of the total page and by the early 2010s this increased to 1/5 or even 1/4 of the page. By the end of the 2010s, while 1/5 or 1/6 of the page appeared most frequently, some images occasionally used 1/3 of the total front page to dramatically present a story. Photojournalism on the page has grown in prominence for the paper, from headline illustration to the star attraction in telling the story, mirroring the Times’ desire to expand into visual reporting over the course of the past two decades.

Below, I compare how three of the most typical and distinct subjects—politicians (55/240), natural disasters (25/240), and war (52/240)—were depicted, offering the starkest contrast between conventional and innovative modes. Through these, I examine how the selected photographs’ aesthetics, role and importance have shifted over time.

Political talking heads

Images of politicians as ‘talking heads’ provide conventional illustration to stories on trade deals, legislation and elections. In Pictures on a Page, Evans discusses the fine line that he feels skilled photojournalists should walk between description and expression:

Good portraits, we have been told, are visual biographies. But if a gifted photographer can attempt a definitive statement there are limits to how far he should try (…) Art may flourish but photojournalism is not well served when the photograph becomes more important than the subject.³⁶

In the early 2000s, photo-editors for The New York Times appear to take this caution to avoid the over-stylised too much to heart—there are numerous examples of this ‘talking heads’ convention on the front pages, with extremely prosaic, literal shots of politicians or public figures mid-speech. Leading figures from the Times have recalled the paper’s photography being regarded in the past as ‘a gray sea of dutiful headshots.’³⁷ It is difficult to imagine that Evans would approve of what he describes as the ‘traditional but boring photograph taken with 55mm lens’ applied near-universally with individual and group portraits of this time.³⁸ A photograph of outgoing President Bill Clinton mid-stroll with European ministers in June 2000 is presented, six months later, in the same visually-flat mid-stride as the staid image (even in hues of brown and grey) of then-incumbent President George W. Bush with his advisors, an image oddly used to illustrate the electrifying headline, ‘BUSH’S LAWYERS ATTACK SUIT BY GORE TO CHALLENGE VOTE.’ [Figure 1] The newspaper’s photography in this period is regarded so clearly as subservient to the text as the image contributes nothing to the reader’s understanding of the story.

By 2010, the shift in the image’s size on the page also lent itself to pointed and unconventional compositions: Elena Kagan’s July 2010 Supreme Court confirmation hearing gives the sympathetic impression of the oppressive and exhausting grilling by lawmakers on the lone individual; a superimposed, projected Hillary Rodham Clinton dominates the screen behind people listening to her March 2011 speech at the Human Rights Council condemning Qaddafi; an August 2013 front page again shows a group shot of people talking in a surveillance committee, but an eye-catching use of negative space provides graphic interest where it lacks as a story-telling device. While the talk-and-walk photography still exists, particularly in the run up to presidential elections, the newspaper demonstrates a greater willingness to experiment with the photograph as a storytelling element—or as a purely aesthetic device.

As previously mentioned, there are instances from the mid to late 2010s where the image serves more as a design element than what Evans describes as a ‘second class’ type of photograph, ‘the descriptive or information photograph.’ These types of images, in their most engaging forms, do ‘not have dramatic action of symbolism. It may at a glance look dull (…) or it may catch the eye easily (…) but the essence is that the descriptive photograph offers secondary as well as primary signals, detailed observation which appeals to the intellect rather than excitement.’³⁹ However, several later photographs featured on the front pages seem to brazenly flout Evans’ caveat that ‘the detail must, of course, be relevant to the observation the picture is making, not just clutter.’⁴⁰ Taking up an entire third of the front page in November 2019, a massive, disorienting image shows the various layers of the House of Representatives during a vote of significant historical importance. Underneath the headline ‘FRACTURED HOUSE BACKS IMPEACHMENT INQUIRY,’ the wide and off-centre shot dwarfs the Speaker of House, Nancy Pelosi, in favour of a dispassionate composition highlighting the numerous, nameless aides behind the benches rather than the tensions between politicians on either side of the aisle. [Figure 2] The votes of each Representative are projected on the back-most wall, but on a busy damask wallpaper—indeed, decoration of all kinds fill up the frame, including neoclassical architectural details. It eschews a more obvious, rousing or symbolic image for one that is more observational behind-the-scenes, striking a contrast between the traditional pomp-and-circumstance of the building and elderly politicians with a line of the young staffers’ modern MacBooks. The story it visually overshadows is a more straight-forward report of the vote split along party lines, as expected. Photographer Erin Schaff’s ostentatious tableau is fascinating—but radically different from the type of utilitarian photography Times readers would have been accustomed to two decades earlier.

Political photography in this period of front-page images appears to become increasingly self-referential regarding the role and presence of the media. However, as scholars like Kiku Adatto have written, the desire to ‘unmask’ the artifice of the photo-op has done little to diminish it.⁴¹ A typical presidential campaign photo—Barack Obama courting the endorsement of New York mayor Michael Bloomberg through a photo-op at a diner in December 2007—is taken in an unexpected angle, on the other side of the glass with a crowd of other photographers muddying the legibility of the star subjects. [Figure 3] The spectacle and manipulative potential of the media is a continuing theme in the newspaper’s front page coverage during the Trump administration, echoing the public’s deepened distrust with institutions and media: a story of an ousted White House staffer is detailed alongside an image in which the subject is in the background, foregrounded by a team of cameras and microphones pointed away from him; a meeting between American and North Korean leaders in July 2019 avoids the conventional mid-range or up-close shot for an oddly bright, washed-out one that pulls far away, revealing stiff, awkward stances in wrinkled suits with the security and camera men, whose lenses angle out towards the viewer (the article describes a staged, ‘made-for-TV spectacle’).⁴²

‘Natural’ disasters

Environmental devastation and the large-scale destruction from hurricanes, earthquakes, wildfires and tsunamis—both internationally and domestically—are instant candidates for the lead image of the front page of The New York Times from 2000 up to 2020. Though such events have always provided spectacle, the images to these news stories are more frequently linked to a larger, systemic narrative—the ongoing climate and ecological crises. Evans frames this coverage through ‘the traditional idea of decisive moment as a news climax’—and certainly these events have a sense of time-based urgency to capture the event in process or its immediate aftermath.⁴³ Tropes throughout the period include human-scaled demarcations of the destruction eliciting emotion through an individual character (a car smashed by fallen buildings, a house’s roof haloed by sinister bellows of smoke and fire, a family escaping from rising tides) as well as images that focus on the raw spectacle of the phenomenon itself (a dramatic plume of smoke from an erupting volcano in Japan provides a dynamic vertical image in April 2000, but no sense of the damage it has caused).

Stylistically, the images of natural disasters veer between documentary images of the people impacted and more abstract or stylised images portraying the ferocity of nature, both to differing effect. Relating to the aftermath of Hurricane Katrina in September 2005, a line of people walking along railroad tracks has no dynamic angle or visual pattern—instead the crowd is framed in an impersonal, faux-objective documentary style not unlike the depiction of refugees.⁴⁴ [Figure 4] Images of environmental chaos later in the 2010s appear less documentary in style, focusing more on details of visual interest and strange or amusing juxtapositions. In September 2012, the featured image relating to a story of another hurricane that hit Louisiana takes on a more surreal or humorous approach. With a tone similar to an idiosyncratic Martin Parr photography, the image shows a woman in a sundress calmly wading through the knee-high water with a clueless looking dog in a cage. [Figure 5] Others, like an October 2010 story relating to flooding, again present self-referential images that show back-stage representations of media. Images related to stories about long-term or systemic environmental damage, such as pollution and climate change, tend to employ more abstract and visually experimental strategies: a sickly hazy, yellow-hued image with a blurry cyclist in China heads an article about the 2015 Paris Talks on greenhouse gases. [Figure 6] A focus on striking patterns with blocks of colour appears in other, more illustrative scenes such as the crags of fractured white walls framing fleeing students after a quake in Indonesia in 2009. The use of more creative and abstract techniques appears to correlate with the contemporary perception of climate change and environmental disasters as systemic and delocalised.

Although many of the aforementioned images on the front pages of 2000 to 2020 give the visceral impression of ‘being there,’ another aesthetic grows in prominence that quite literally diverges from the ‘on the ground’ approach. The popularity and prominence of drone photography in telling environmental stories has continued to grow in this period since their popularisation in the early 2010s. A 2014 article in British newspaper The Telegraph welcomes the reader to ‘the brave new world of drone photography’—a medium it describes as ‘in its infancy’—and includes comments from the newspaper’s ‘dedicated drone photographer’ who was hired in late 2013.⁴⁵ In 2018, The New York Times published an article regarding the types of drones and other equipment its staff photographers rely on, describing how drones have improved substantially across the decade, even becoming a tool for reports on drought, fire damage, and migration patterns. Staff photographer Josh Haner enthuses that ‘drones have democratized aerial imagery, as it no longer costs thousands of dollars to rent a helicopter or a plane to make images from above.’⁴⁶ Satellite and drone photography have particularly been applied to environmentally focused stories: a September 2017 article shows a before and after of a submerged neighbourhood in Texas working effectively with the headline ‘Disaster With “No Boundaries” Devastates Rich and Poor Alike’; an odd perspective from a drone shows an abstracted landscape including a destroyed levee for a story about severe flooding in the Midwest in April 2019. [Figure 7] Though contained images of disasters are still produced, particularly within the ‘spot news’ genre, there appears to be a rise in more symbolic and abstract photography that attempts to portray the less-perceptible systemic links between the ‘local’ and ‘global,’ linked to a larger climate change narrative. Visual reporting increasingly over this period frames individual natural disasters as part of a manmade and globally enmeshed crisis.

Contemporary conflict

From the early 2000s, during the beginning of the wars in Afghanistan and Iraq, the bombast and schmaltz of heavily symbolic images take precedence in the Times’ front pages. As photojournalists struggled to take interesting and eye-catching images from an embedded position (with heavy censorship by the US military enforced later in the Iraq War),⁴⁷⁴⁷ more conceptual strategies such as ‘aftermath’ centred images begin to appear in news photographs. Much discourse already exists regarding visual qualities and tropes of photographs showing the violence and displacement of war.⁴⁸ In her 2010 book About to Die, Barbie Zelizer has recounted the ways in which journalists crafted a narrative—showing American soldiers playing with local children in Iraq and Afghanistan—and upheld an embargo for many years on depicting dead or injured American soldiers while publishing images of their belligerents’ corpses. Criticism around 2004-5 of the American media’s squeamishness about showing the violence of war led to a particular moment of introspection for news outlets.⁴⁹ The ethics of looking, and of not looking, were front and centre in media discourse—and, at a time when photographers began to fray under the constraints of embedded journalism. In concluding remarks updated in 1997, Evans contends there are four questions to help the editor decide whether or not to publish an ‘offensive’ photograph:

Is the event it portrays of such social or historical significance that the shock is justified? Is the objectionable detail necessary for a proper understanding of the event? Does the subject freely consent? Is the photograph expressive of humanity? Not all these questions need to be answered in the affirmative, but at least one must to justify a shocking publication.⁵⁰

Evans follows his criteria with an example of Kenneth Jarecke’s controversial Gulf War image—the body of an Iraqi soldier who had burned to death—as a necessary image for understanding the event.

Within the same criteria, The New York Times coverage of conflict is fairly conservative in this regard, yet there was visible change in how it featured on the front page. Particularly in 2002 and 2003, soldiers are shown as departing heroes, with tearful goodbyes (February 2003), [Figure 8] waiting together in anticipation (December 2002), or a show of strength in numbers (March and April 2003). A moody landscape by acclaimed photographer Yannis Behrakis features an impressive plume of smoke along a strong horizon in November 2001, accompanied by a caption that reveals ‘The Air Force has begun using B-52s against the front lines of Taliban troops. These bombs from a B-52 hit about 15 miles north of Kabul yesterday.’ Though hardly the ‘shock and awe’ seen later in the Iraq War, the image emphasises a display of force in of itself rather than the destruction these weapons cause. A sanitised version of war is generally present (e.g., March 2002 Israeli raid) for much of this period—injured but living victims feature on an August 2002 article about a bombing of a Jerusalem university campus, but death itself seems limited to the Iraqis with images such as the nondescript bundles of dead protestors’ bodies in May 2003 and the procession of a funeral with an empty coffin for Ayatollah Muhammad Bakr al-Hakim in Baghdad in September 2003. These images show crowds of people in grief or outrage, rather than focus on individual or personal moments of emotion as seen with the coverage of American soldiers.

Around the same time as the publishing of the shocking Abu Ghraib pictures in April 2004 (famously on television and in print by CBS News and The New Yorker, respectively), there is a shift in what is shown with a growing ambivalence to American military in the paper and more individualised portrayals of Iraqis.⁵¹ Perhaps the newspaper’s editors, like the editors of the wire agency Agence-France Presse that Gürsel describes, ‘bemoan[ed] loss of control over powerful images’ and sought to rectify a long-standing reliance on symbolism to convey information.⁵² Bodies now appear on the front page: victims recognisably and unnaturally slung over a bombed bus in Israel in September 2004, burnt corpses of Blackwater mercenaries hung along a bridge over the Euphrates in April 2004, [Figure 9] a tender close-up of a mother holding up a picture of her missing son outside Abu Ghraib prison in June 2004.⁵³ The perspective of these images brings the camera right up to the action, giving the viewer the feeling of being in the crowds. Though offering a different perspective, these images still present war through a reportage style.

Other images in the mid-2000s take another tactic, showing conflict through what has been characterised as ‘Late War Photography’—photographing ‘after’ the event has happened, usually with an emphasis on landscape and classical compositions.⁵⁴ Following the end of the July War in Lebanon, a September 2006 photograph shows its impact through a striking line of trucks filled to the brim with rubble leaving from the city to the countryside—a billboard for a tourist attraction captured on the right side of the photograph draws a wry comparison between the pristine natural rock formation and the manmade one being formed. [Figure 11] Conflict is articulated through the landscape rather than pathos-driven images of people in distress or the spectacle of destruction.

By the mid to late 2010s, it is no longer surprising to see images of the dead for accompanying stories about conflict, such as a small body of a child in the August 2015 story ‘Firebomb Kills Palestinian Toddler in West Bank’ or even more graphically in a May 2018 article (‘A Documenter of Afghan Victims Becomes One’) which draws an eerie contrast between the dead man’s face and the wreathed vibrant portrait above him. [Figure 10] Rather than focusing on the grief or anger of bystanders, these desaturated, sombre photographs present the body as the focal point of the image. Equally, photographs of the period show wounded, less dignified images of American soldiers (July 2015) and graphic images of wounded and dead civilians immediately after attacks (June 2017).

In contrast to the gritty, mid-2000s reportage, there are images in the later 2010s depicting the consequences of violence and war—detainees in a military prison (September 2014), disrupted domestic lives (May 2015, Yemen; September 2018, Iraq), refugees fleeing conflict (October 2015, July 2016)—which are demonstrably more elegant. They are beautiful and geometric images in black-and-white, luminous and dappled painterly portraits, and eye-catching but actionless moments of emotional drama rather than dynamic, grand events. Instead, these are quiet moments within extraordinary circumstances—individuals surviving within carceral infrastructure or propelled into new lives.

Only a small number of photographs from this dataset of front pages were explicitly created by amateurs (2/240), reflecting the Times’ recent history of using amateur images to tell exceptional stories where professional photographers lack access—such as in conflict stories.⁵⁵ Above a July 2008 article about Al Qaeda, a bright green, blurred shot provided by the terrorist group reveals a new role for the Times not as an image-creator but as curator, authenticator. The caption for it reads: ‘Abdelmalek Droukdal, third from right, an explosives expert and leader of one of the most potent Qaeda affiliates. The photograph’s authenticity was verified by The New York Times.’ The ethics of reprinting an image provided by a terrorist group aside, the amateur photograph is used as a more traditional or conservative editorial image. It is a document, an artefact as evidence. The use of amateur material around this period grew in prominence with footage provided by amateurs to the newspapers; following the Arab Spring of 2011, a quickly deteriorating situation in Syria resulted in images of the conflict being ‘curated’ rather than ‘covered’ by The New York Times. The paper endeavoured to authenticate amateur photographs and videos in lieu of on-ground staff photographers for this more ‘hard news’ content.⁵⁶

An ‘Image-First’ Future

In 2018, The New York Times Company announced, with some fanfare, a monumental undertaking: in partnership with Google, the 167-year-old newspaper planned to digitise its photography archive of approximately 6 million photographs stored in its famous basement archival ‘morgue.’⁵⁷ This gargantuan effort would rely on Google Cloud storage capacity and the platform’s ability to process metadata from both the writing on the back of each photo as well as by recognising objects and locations through targeted APIs (application programme interfaces).⁵⁸ Through this automated process, massive amounts of metadata can be generated without time-consuming manual, human input. Who is this initiative for, though? A decade prior, Google had unveiled a similar project with TimeWarner for the digitalisation and hosting of LIFE magazine’s prestigious photography archive. Though the 2008 website is a far cry from the algorithmic image analysis that the current Times partnership bodes, the LIFE project’s dated design and database stands out in another crucial way as Web 2.0: it was heralded as a public resource boasting millions of images.⁵⁹

Unlike Google’s work with Life, archival digitalisation for The New York Times is not intended as a public resource. Herein lies a key issue with press archives held in company hands in Web 3.0: the continued privatisation and financialization of culture, with little incentive to open up valuable resources to public scrutiny. Recent controversies with Magnum Photos’ own mismanaged archive of press and documentary images in 2020 for instance resulted, thus far, neither in a measured re-evaluation of the archive nor in a clear strategy for improving keyword terms and related metadata, but in a removal of access to the database altogether.⁶⁰ For The New York Times, despite having digitised over a million images, these photographs are on an internal database.⁶¹ On the newspaper’s image licensing portal, a conservative ‘4,000 historic photographs’ is quoted instead.⁶² With the lack of a formal, public-facing archivist, a researcher might have better luck visiting the Times’ online store and requesting a ‘custom photo print’ for between $75 and $410, as this prompts the buyer to enter image location details.⁶³ Naturally, the discrepancy between internal and commercial access to images has much to do with copyright: ‘Images credited to wire services, private archives, museums, individual photographers and other photo agencies are not available.’⁶⁴

My concern as a researcher, though, is not with the commercial archive but the public one. The inconsistent public outlet for the photo digitalisation project is the Times’ curated ‘Past Tense’ page, which acts as a depository listing recent articles that make use of the archive as well as more tailored discoveries from the team scanning the images themselves.⁶⁵ More information on the internal photo archive is available publicly from the Times’ developer and designer blog, ‘Open,’ where case studies of the trial and errors of machine learning to create a useful CMS provide insight into the newspaper’s use of the archive: primarily as a tool for its photo editors to create ‘a richer, more visual report for our readers.’⁶⁶

Conclusion

This desire for ‘more visual’ content hints at the broader purpose of this initiative, and to the changes at The New York Times as a whole—a move towards image-led reporting.⁶⁷ As my analysis has demonstrated, the Times has transitioned over the course of the past two decades from its stuffy and old-fashioned predilection for the image as subservient to text, to images engulfing more and more of the digital and physical front page. The national edition of the newspaper has increasingly incorporated visual experimentation in its use of photography, employing formalist, ‘artistic’ visual strategies for what were once more conventional editorial subjects. Additionally, I believe that this aesthetic shift is linked to an institutional trend, that most contributing photojournalists to the Times are now freelancers. Freelancers, by economic necessity, pitch images and stories to multiple sources such as publications, corporate sponsors, NGOs, and galleries as well as sharing directly with their audience through social media channels. The aesthetic trends explored here, then, are a further indication of the hybridisation of photojournalism with other genres, such as fine art and commercial photography, to adapt to multiple platforms, purposes and audiences. The long-held hierarchy between text and image has been disrupted, with images taking the lead.

However, through analysing the content and character of this dataset I found that the company’s ‘image-first’ approach is not reflected in its digital archive. Indeed, many of the same tensions from its focus on ‘Page One’ arise: print versus digital, text versus image, ephemerality versus permanence. Firstly, much of the material readily available to researchers is poor-quality scans from microfiche—with near indecipherable, high-contrast black-and-white representation of illustrations for the sake of more legible text. Secondly, while colour PDFs and JPEGs have become available in recent years, the lack of precision regarding early and late editions can potentially erase entire images from the digital archive. Finally, as discussed, the ‘archives’ of large news organisations exist explicitly to serve commercial interests with little incentive to open up their company to the scrutiny of researchers. Investment in better archival preservation coupled with open access would undoubtedly transform their potential for research; this could better enable interdisciplinary collaboration between art-historical researchers like myself and data scientists, combining the depth of aesthetic sensitivity with the breadth of machine-learning.

The current limitations are best illustrated in one final example. Prior to manually collecting information and images to form the dataset examined here, I had been working in a contemporary art organisation. Whilst there, I happened to work on an exhibition for an artist, Rayyane Tabet, who tended to use found materials for deeply personal conceptual works. One such piece was his Friday, September 1, 2006 (2006–2019), a found newspaper (in his words, ‘stolen’) from when he was studying in New York. [Figure 11] The image on the front page of The New York Times had caught his eye: ‘The caption read: “A River of Rubble From the War in Lebanon”.’68 The framed newspaper, Tabet’s found artwork, shows Dimitri Messinis’ photograph, discussed previously as within a trend for so-called Late War Photography. To my dismay and confusion, this New York, Late Edition of the newspaper has completely evaporated from the digital archive. On both ‘Today’s Page’ and ProQuest’s archive, the photograph, in exactly the same spot, was instead a sports photograph of Andre Agassi at the US Open. In the slippage of digital archives, Messinis’ image has disappeared from the front page entirely. The photograph only exists as a small, oddly cropped insert into a preserved, related online article.⁶⁸ While the Times may see digitalisation as the road to an image-first future, its inaccessible and inconsistent archives keep everyday photography as ephemeral to the public and researchers as it has ever been.