9. Images, Art & Video - oritenting

[JunsolKim]

Post questions here for this week's oritenting readings: Collins, Randall. 2009. “The Micro-sociology of Violent Confrontations” and “Confrontational Tension and Incompetent Violence” (beginning of Chapter 2) from Violence: A Microsociological Theory: 37-43.

[konratp]

I appreciate the wide variety of sources Collins analyzes in writing his book on violence in society. So far, many of the papers we have read in this class have employed exclusively highly technical quantitative methodologies, but not very many of the papers brought together so many different types of sources.

I do wonder, as photos and videos -- especially of violent situations, as we can currently observe in Ukraine -- are often taken with by someone with a personal investment in the situation and hardly completely neutral, do researchers run a higher risk of mistaking subjective perspectives on violent situations as objective truths because the pictures/videos are "hard evidence?" How have the standards surrounding the usage of such multimedia sources evolved as deep fakes have become more prevalent in recent years?

[facundosuenzo]

Collins mentions that "Methodological purity is a big stumbling block to understanding, particularly for something as hard to get at as violence." and I wondered to what extent, as junior scholars, it is recommendable to deeply specialize in a particular methodology (ex. ethnography, survey-methods, computational methods, you name it) or, instead, bolster our research in enriched mix-methods strategies? A dissertation could be an excellent object to explore abandoning this purity, but what are the hidden costs -if any- in embracing methods that sometimes are ontologically opposite? Losing rigor?

[GabeNicholson]

Thought this was a great read and taught me a lot of things I had no idea about. For instance, most soldiers don't even fire their weapons and the majority of war-based fighting tactics are trying to get soldiers to not be afraid.

I wonder what the potential can be for future deep learning algorithms that will be able to predict fights ahead of time based on body posture and other language cues? Also, what kind of knowledge does studying violence offer in terms of public policy and policing strategies? I know the authors briefly mention that war tactics have been improved due to social science advice in the past for getting soldiers to conform—but has any transformational knowledge like that happened recently in our society that can be credited to the sociological study on violence?

[NaiyuJ]

Some thoughts after reading this:

In the toolkit of every professional photographer is a collection of lenses – some for close-up shots and some for wide-angle shots.  The same applies to social scientists.  Doing personal interviews and having detailed, in-person conversations with historical participants in an event at one locale, allows the scholar to have an in-depth, personalized understanding of the matter; but, in absence of other methodological tools, one always has to wonder which parts of the findings are generalizable, universal, causal facts, and which parts of the findings are local, particular, and accidental facts.   In X’s analysis, we have presented the factors, details, and stories of one place at one time.  We cannot, however, use these “close-up shot” data to distinguish the local and transient factors from the global and fundamental ones.  Textual and Image analysis (the analysis of photographs of conflicts and violent situations in this case) is the fish-eye wide-angle shots that give us the complementary perspective.

[isaduan]

As intense data collection becomes available and we are increasingly capable of reconstructing the situations (never perfect, but ever better!), I wonder whether the purpose of social sciences is also shifting, from understand the 'absolute objective truth', to understand each person's perspective, why he, she, they behave the way they do from their perspective; from what it's easy and convenient to talk about, to what really matters to living individuals ... to that end I am optimistic, and think social science could be a kind of immersive technology invoking sympathy and connections between different individuals, groups, and cultures.

[ValAlvernUChic]

The chapters got me thinking about the mechanisms of violence and how our perception of what it is and entails is a function of many things outside reality. He refers to violence as a ritual and he seems to, at least to me, imply that "symbolic violence" as Bourdieu terms is not real violence. With technology being so advanced, such that digital terrorism, cyber attacks, cyber-bullying have all become legitimate forms of harm, I can't imagine that this should still(if it ever) hold true. On his methodological stance, I completely agree though - it is often easy to miss the forest for the trees when it comes to restricting ourself to a single method to make sense of the world.

[sudhamshow]

I am not entirely sure about the fear and non-performance arising due to confrontational tension that the author alludes to. I remember an interview with a reported and a teen from the suburbs in Chicago who had been detained for gun violence. The teen said he wasn't afraid of losing his life and had very little to hope for. When people have little to look forward to, I think they don't fear losing their lives. The interviewer also discussed the epidemic of teenage gun violence, how they found a fraternity among gang members and how more people joined the streets who were not afraid to pull the trigger.

[Jiayu-Kang]

The chapters got me thinking about the mechanisms of violence and how our perception of what it is and entails is a function of many things outside reality. He refers to violence as a ritual and he seems to, at least to me, imply that "symbolic violence" as Bourdieu terms is not real violence. With technology being so advanced, such that digital terrorism, cyber attacks, cyber-bullying have all become legitimate forms of harm, I can't imagine that this should still(if it ever) hold true. On his methodological stance, I completely agree though - it is often easy to miss the forest for the trees when it comes to restricting ourself to a single method to make sense of the world.

^^ Adding to your thoughts about the mechanism and perception of violence, I'm also thinking about the rationale behind differentiating "symbolic violence" and the "real violence." Although I like the micro-situational approaches to understand violence, I'm not fully convinced by how the author dismissed the macro-cultural approaches, since it seems to me that there are more connections than distinctions between the two.

[pranathiiyer]

The chapters got me thinking about the mechanisms of violence and how our perception of what it is and entails is a function of many things outside reality. He refers to violence as a ritual and he seems to, at least to me, imply that "symbolic violence" as Bourdieu terms is not real violence. With technology being so advanced, such that digital terrorism, cyber attacks, cyber-bullying have all become legitimate forms of harm, I can't imagine that this should still(if it ever) hold true. On his methodological stance, I completely agree though - it is often easy to miss the forest for the trees when it comes to restricting ourself to a single method to make sense of the world.

Adding onto this, with the advent of VR and video games that simulate very real experiences of fear and violence through completely vicarious experiences, I wonder how these mechanisms and perceptions of violence might change?

[chentian418]

The authors mentions that technological improvements hypothetically might reduce bystander hits, while the experience of early first-twenty century wars continues to show problems of bystander hits and friendly fire. These pattern suggest that the root problem is not the technology, but the fog or tension of combat itself. Therefore, I am curious that, how do technologies like remote radar, satellite sensing and highly accurate navigation systems distinguish violence from different similar scenes? And will those technologies mislead humans with problems like misclassification?

[Qiuyu-Li]

This is the first material I read in criminology and it’s very interesting! I can imagine that videos can be a significant tool for micro-sociology criminologists to study the situations of violence, and I’m actually curious about what they are studying here? Of course, we know anger, hatred, and sadness trigger violence, but what can we do about it?

[Hongkai040]

This study reminds me Apple's controversial child sexual abuse material (CSAM) detection. Apple censor pictures and videos uploaded to iCloud to see if they're SCAM related. As far as I know, Apple compares those hash values of those contents to National Center for Missing and Exploited Children's dataset. So basically they only detect CSAM in the dataset. This week's topic has to do video/picture processing. I think it's possible to use supervised/unsupervised machine learning algorithms to detect such videos and pictures. Why do tech companies like Apple don't use them? I feel strange. Does this imply the current limitation(like low accuracy) of using machine learning algorithms to extract information from videos and pictures?

[LuZhang0128]

This is very interesting reading. However, when applying to our own research, I'm thinking how available are the resources? Meanwhile, to what degree can we trust our data? Just like the misinformation label on Twitter for text, violent pictures and videos are oftentimes highly sensitive and will be monitored by social media platforms. Also, just like what has happened recently, there can be tons of fake pictures and information. How can we collect meaningful data on a large scale and prove the validity of the data?

[YileC928]

Very interesting read! As many of my classmates have mentioned, I am also wondering how to evaluate the validity of data, in particular images and text - since they could be produced and interpreted subjectively.

[hshi420]

I'm curious about how would AI generated content should affect the application of video or image analysis in criminology field.

[Emily-fyeh]

Every time when reading this type of material, I would always sense that morality and ethics are changing at a fast pace than in the past. And for now multi-national enterprises, giant tech companies have the power to decide the process of content review/report and to decide if the online content abides by the community standard. And we have seen this power prevails and outshine the traditional jurisdiction, creating an obscure effect on information wars.

[hsinkengling]

As the author suggests, society seems to have its own bias for what violence should look like, I wonder if that kind of bias will also reflect in the kind of data one can gather once we're actually doing the quantitative analyses of videos of violence. People might be more inclined to upload heroic rather than confrontational/fearful footage of fighting.

[ttsujikawa]

As many of classmates mention, value of violence is increasingly becoming more subjective than before and it is getting harder to recognize distinguish what is meant to be 'violence' (even for human). In this transition, violence could be highly contextual in any forms. Then, how could we apply this dynamics into machine learning model?

[sizhenf]

Very Interesting read! I'm curious as to how computational algorithms can recognize violence with different situational components.

[melody1126]

In this chapter, the author points to the rich data sources through which we can study and theorize about violence. I can trace the abduction here – that the author proposed to an investigation into something as abstract as violence using videos and images is very interesting and shows a case for not just sociological inquiry, but also one that requires knowledge of psychology into "confrontational tension and fear" to make meaningful interpretations of the image and video sources.