Babies in bathtubs: Public views of private behaviors represented in the Flickr domain

Authors


Abstract

Publishing personal content online raises issues of openness, access and privacy. While these issues have been treated in the literature, a substantial subset is youth-oriented, typically focusing on the potential for negative outcomes from such disclosure and youths' presumed lack of awareness of the consequences of their publication activities. Recent research suggests, however, that adults also demonstrate a lack of discretion in their own publication activities. Through examining a particular class of content - adult publication of photos depicting “babies in bathtubs,” and other candid and bare moments, to Flickr, a Web-based photo sharing and management application - this study aims to quantify and describe issues of openness, access and privacy to raise awareness that these issues are not limited to generational distinctions, by 1) characterizing images across multiple topical domains from two contextual perspectives, contributor and user, and 2) identifying associations between these image characteristics and contributor behaviors and profiles. The primary assumption motivating this research exercise is that online indiscretion is a youth problem, but in fact, it is broader than that; the collective “we” are all subject to failing to meet the assumed societal norms for deliberate or accidental privacy and discretion when publishing personal content to the Web.

Introduction

The populist emergence of the read/write Web has enabled everyone and anyone to be a publisher. Flickr, an online photo sharing service, is one example of this emergence of user-generated content and the potential, depending on user preferences, for universal access and subsequent loss of transparency between private and public lives. Via Flickr, the personal archive is made available online, with the effect that the family photo album, once confined to living rooms, is brought into the equivalent of the town square. A segment of recent literature on privacy and the use of social technologies by youth, both in academic and popular discourse, has been from the admonitory perspective of parents and concerned others. It has focused on the potential negative outcomes from online, public disclosure of private information, including personal identifiers and potentially embarrassing photos, and the presumed lack of youths' awareness of the persistence and discoverability of these elements. But other recent revelations in research and the press have shown that many adults are perhaps not as aware of these professed privacy concerns themselves and, in fact, frequently post and disclose photos and information that may result in unintended public disclosure. A recent study by the Pew Internet and American Life Project found that 39% of teens restrict access to photos posted online most of the time compared to 34% of adults; 38%-24% of adults restrict access sometimes; and 21%-39% never. (Lenhart et al., 2007).

The research investigation presented in this poster examined a particular class of content in an attempt to quantify and describe issues of openness, access and privacy from adult publication of online content – photos posted to Flickr depicting “babies in bathtubs” and other various stages of undress, or what some particularly zealous prosecutors might classify as child pornography. The intent of this research is not to be lewd or moralistically subjective, but to highlight that issues of privacy and identity disclosure online are not just oriented to youth and young adults, but is an everyone condition; it's not just Junior posting questionable content, but Mom and Pop too.

Literature: Social Technology and Privacy

Much work has been done in recent years regarding perceptions and uses of privacy measures online, primarily focusing on youth populations and users of online social networking services (SNSes). Palen and Dourish (2003) provide a general model for the complex issues involving privacy and disclosure in online mediated environments. Their observation regarding the disruptive role of new capabilities in this process is particularly useful: “Technology itself does not directly support or interfere with personal privacy; rather it destabilizes the delicate and complex web of regulatory practices” (Palen & Dourish, 2003, para. 37). Gross and Acquisti (2005) in their study of Facebook noted the weakly deterministic role of software in disclosure of personal information. In a follow-up analysis, Acquisti and Gross (2006) found that Facebook users report high levels of concern over privacy and related issues but by and large do not implement privacy controls that are available to them.

Barnes' (2006) analysis feeds into the common misconceptions regarding online disclosure, asserting that teens putting information online is a “problem” of which they are insufficiently aware. This scare-tactic analysis repeatedly mentions the hype over the danger of “online predators,” using two newspaper articles reporting that “several young girls have been molested by men they have met on social networking sites” in support of this position (Barnes, 2006, para. 17). By contrast, boyd (2007) lays out a framework for understanding how youth actually understand and use SNSes, and how adults (especially educators) should interface with them in these spaces. boyd provides a further rigorous and refreshing corrective to Barnes' hysteria with some of the initial results of her two-year ethnography of teen practices in online social spaces, framed by her development of the idea of a “networked public” and definition thereof.

Lenhart and Madden (2007) deliver results from a measured and wide-ranging survey of American teenagers on attitudes towards and practices relating to privacy online and with regard to SNSes, offering a refreshingly straightforward and respectful picture of teens. On the whole, “…many youth actively manage their personal information as they perform a balancing act between keeping some important pieces of information confined to their network of trusted friends and, at the same time, participating in a new, exciting process of creating content for their profiles and making new friends” (Lenhart & Madden, 2007, pp. i-ii). Teens, they say, “consistently say that the decisions they make about disclosing personal information on social networks and in offline situations depend heavily on the context of that exchange” (Lenhart & Madden, 2007, p. 2). In further research, Lenhart et al. (2007) also note that “two-thirds (66%) of teens with an online profile say they restrict access to it in some way, while just 50% of online adults with profiles restrict access” (p. 13).

While it's clear that the framing of privacy issues and teen identity online has been heavily influenced by popular media narratives, it has not been subsumed by them. We would argue that excessive focus on issues of and concerns with privacy are obscuring more basic issues at play in expressions of online identity - up to and including those conceptual issues surrounding how contemporary users actually conceive of and define “privacy.”

Problem Space

In a widely-publicized case from 1999, a technician at a Fuji Film processing lab in Oberlin, OH turned over to law enforcement a roll of film belonging to Cynthia Stewart and containing pictures of Cynthia's 8-year-old daughter, Nora, at various stages of taking a bath. The case was resolved after a long legal battle, but was one of a spate of similar cases from the late '90s. The world has changed since then and, for many people the idea of taking their pictures to a photo lab for processing is as foreign as sitting for a daguerreotype. Instead of risking the nosiness of a photo technician, possibly in part due to cases like Stewart's, they just put their kids' naked pictures up on Flickr, where only they and their friends can see - right?

As a recent Washington Post article detailed, Meredith Massey uploaded three pictures of her children skinny-dipping to Flickr and marked those untitled and unclothed pictures “private,” restricting access to her parents only. She was surprised when she subsequently discovered many had been viewed thousands of times (Marr, 2008). Flickr is investigating the breach. Her story depicts a revealing assumption that users of online services like Flickr may have: an assumed or delineated community of readers. Publishing content to the web takes place with the default assumption that the content is intended to be public and that extra steps by contributors, such as selection of authorization controls, is necessary for exerting more control over published content, and two, that taking such control, as Massey reportedly did, does not necessarily translate to sufficient enforcement, and circumvention, however unintended, is possible. A further implication is that these decisions for openness, access, and privacy are made by the contributor and not necessarily the subjects. Flickr places parameters on users aged 12 and younger, requiring parental permission for the creation of a “family account.” So, while others, including children, may be the subjects of the contributed photos, these others are not necessarily participatory in making decisions on who views and uses their depictions. Preliminary search activities show this to be true for both those subjects assumed to be related (e.g., parent/child), and those that appear to have only a time-in-place oriented relationship (e.g., tourist photo).

Objectives

The primary assumption motivating this research exercise is that online indiscretion is viewed as a youth problem, but in fact it is broader-based than that. The objectives of this study are to raise awareness that issues of openness, access, and privacy are not limited to generational distinctions. Rather, the study is undertaken to stamp out the sentiment that youth are primarily unaware, indifferent, or unresponsive to implications of posting content; in fact, the collective “we” are all subject to failing to meet the assumed societal norms for deliberate or accidental privacy and discretion. On the contrary, youth, as digital natives, may have better innate senses about the appropriateness (or not) of disclosure, something which recent research seems to suggest (Lenhart et al., 2007).

Proposed Methodology

This study is an image-centric analysis within the Flickr domain. Our data consists of a set of images uploaded to Flickr, a photo sharing, online social utility. Preliminary searches taken in preparation of this study proposal yield result sets ranging in the hundreds to tens of thousands, Data collection for our study will be enhanced by utilizing the Flickr API to collect a results set from a string of search terms compiled for image retrieval across both the full-text fields and tags. From the set of images returned resulting from use of the Flickr API, the first 500 images returned will be considered for analysis. From this result set, we will characterize images in the aggregate across multiple topical domains from two contextual perspectives, contributor and user: image type (e.g., digital camera capture, scan); subject orientation (e.g., family association, other association); level of depiction (e.g., truncated, full frame); access (e.g., application of access restrictions); use (e.g., application of use licenses; number of views); description (e.g., tagging and annotation); and distribution (e.g., group memberships). Further, we will look to identify associations between these image characteristics and contributor behaviors and profiles: identification (e.g., real names and pseudonyms); subscription (e.g., free or fee-based service subscription); and community (e.g., contacts). The study design is currently ongoing. Projected date for completion of data collection is June 2008, with analysis to be shared in the poster display at ASIS&T 2008.