Title: Curating Death In American Horror Films
For the first time ever, this project will curate deaths in the first century of American horror films using an online network of international, amateur, volunteer archivists to extract specific data characterizing cinematic death events: demographics (to include headshots), time, location, technology used, type of death (loss of body and/or loss of identity), and box office. While headshots will be crafted into static and dynamic digital art, the online database will be accessible to scholars and amateurs alike with project-specific analytical tools to visualize the data using a variety of charts, tables, and maps.
PART 1 (Oct 6): NARRATIVE (Humanities Questions addressed)
Using the project’s online visualization capabilities, scholars will be able to consider the following questions (and many others) based on a variety of themes suggested here:
1) Structure How soon does the first death occur and when do following deaths occur with respect to film runtime? In other words, what’s the “death rhythm” and how has it changed (or not) over time? Given that Hollywood tends to follow filmmaking formulas, is there a formula for deaths and is it reflected in bigger box office receipts? Further, considering Adorno’s “Culture Industry,” what does this desensitization of and expectation for death in entertainment do to spectators and their acceptance of deaths caused by government and industry (i.e., wars, pollution, deadly products, and health care limitations) especially in the 2nd half of the 20th century?
2) Race When do minority deaths begin to appear and how does their prevalence change over time? What role does “having a name” play in the deaths of minorities? Do they have: no names, first names only, or first and last names? How does this differ from whites? How does this named identity critique/affirm the government treatment of minorities over time? Did these films lead or lag the establishment of minorities as equal citizens with named identities and how did they affect white spectators’ acceptance of minorities?
3) Class Who are those “privileged” with onscreen deaths based on race, gender, age, and class? While there is a given tension between being privileged with death and the deprecation of dying, inclusion in the film is the more important metric. Who do the white, middle-class spectators see dying more often: themselves, those above them (economically and/or socially), or the Other? Which cohort of death does better at the box office? What is the role of Freud’s sadism/masochism on the spectator and how does it reflect the “trapped” nature of modern life?
4) Gender How are deaths gendered, specifically considering 2nd Wave feminism and the public/private binary? Where do women die in these films: inside the privacy of the home or outside in the public sphere? What role does patriarchy play in locating these deaths; i.e., is the private home safer than public spaces? How does the ratio of female-to-male deaths change over time and how does it correlate with the women’s liberation movement? Did it change after the rise of conservatism in the 1980s?
5) Modernity Geographically, where do these deaths occur from an urban/rural binary? One typical escape from the modernity associated with an urban setting is the potential utopia of rural life where modernity and its highly regulated and efficient use of time is supposedly absent. How do these films critique/affirm this utopian ideal? Do these films do the work of industry by showing that horrific deaths are inevitable even in the country, so spectators (workers) should just stay put and submit to modernity?
6) Technology What is the role of technology in the death events? What specific technology is used to kill and what level of intimacy is involved? The spectrum can go from zero tech (hands) and ancient tech (knifes/blunt objects), which require a close kill, to more advanced tech which can kill at a distance (guns, bombs, monsters, radiation). Is there a pattern in the technology that correlates with as increasingly technological society? Is there a consistency with knife kills (ancient tech) over time that reflects an impulse for spectators to maintain a level of consistency in their lives? Do the high-tech kills in film numb spectators to the inevitable deaths from new technology that is concomitant with unbridled modernity?
7) Identity When does the “death” of an individual via a loss of identity rather than a loss of bodily life first occur and how prevalent is it over time? This type of death occurs when the body is more or less intact, but the mind has been compromised rendering the prior individual effectively dead. How does this type of death reflect the concerns of humans feeling trapped in a culture that molds them into something they’re not? Where do these specific deaths occur using the urban/rural binary, and how do they reflect on the ability of the individual to escape modernity? Do these bodies with a new identity ultimately suffer a second, physical death and how are spectators expected to respond?
NOTE: This is not an all-inclusive list of potential themes or questions that may result from this project. Given the array of variables/binaries considered, scholars can investigate on the order of 100 possible correlations where just a few correlations may be sufficient data for a publication and add to our understanding of American culture.
PART 2 (Oct 13): NARRATIVE (Environmental Scan)
The work proposed in this project is closely related to cinemetrics, a newer area of film analysis developed by Frederick Brodbeck (http://www.fredericbrodbeck.de/) in 2011 which creates 3-D “fingerprints” (a 2-D image that changes with time) of films by focusing exclusively on its objective structural elements without much concern for its narrative. The film itself is seen as the source of raw data where such elements as color, movement, shot duration, shot frequency, and other “cinemetrics” are used to create a dynamic visualization of a film: its fingerprint.
Brodbeck’s process of creating a film’s fingerprint with dynamic examples is here https://vimeo.com/26584083 and a more detailed discussion of his work is here http://cinemetrics.fredericbrodbeck.de/ He developed specific computer codes (tools) to automate the work to perform such labor-intensive tasks as cut detection, color analysis, and statistical visualizations. He made his software open source for other scholars, and his python scripts and batch files may be found here (note that they haven’t been updated for 9 years): https://github.com/freder/cinemetrics
While Brodbeck’s work shows how a purely objective analysis of film metrics may be automated, the proposed project relies on the spectator’s assessment of the film to capture other metrics that don’t lend themselves to automation (who died, demographics, location of death, method of death, etc.). In this case, the admittedly labor-intensive approach of Yuri Tsivian via the Cinemetrics website is more applicable (http://www.cinemetrics.lv/). Here, spectators are collecting data on film cuts and either tagging the scenes in the “advanced” mode or ignoring tags in the “simple” mode. They rely on the Internet Movie Database website (https://www.imdb.com/) for standardization of movie titles, director, and other information. Spectators may collect and submit their data via an online Flash app (which does not store data locally) or with an app installed on their computer (http://www.cinemetrics.lv/cinhelp.php), which appears to be last updated in 2012 based on the YouTube tutorial (https://www.youtube.com/watch?v=tUAdES1gWgM). Once the data from a film is collected, it is then uploaded to the cinemetrics.lv website.
An example of the analysis possible with this simple data is shown in Barry Salt’s Moving Into Pictures book (may be downloaded at http://www.starword.com/). An excerpt is shown here http://www.cinemetrics.lv/distrib.html where histograms reveal the hidden pattern of ASL (average shot length) in a variety of films. It is this “revealing of hidden patterns” using simple data that motivates the present project in the deaths in American Horror films: what will various visualizations reveal about our culture and biases?
Since mapping will be a primary deliverable of the proposed project, a cinema-specific mapping project called “Mapping Movies” (https://www.mappingmovies.com/) includes relevant visualization elements. Its ambitious project of mapping of movie theaters throughout the world includes interactive maps with a variety of layers (https://www.mappingmovies.com/interactive-maps/), deep mapping, walking tours, and visualizations. Often text and graphics are used to highlight and explain a particularly significant finding from the research rather than hope the user stumbles upon it while interacting with the website. The proposed project will use a similar bifurcated approach: 1) a freestyle, open-ended, dynamic experience for user exploration and discovery, and 2) a selection of curated data with scholarly analysis so users may reflect on significant findings that will inform their future exploration of the database.
While the term “cinemetrics” rightfully conjures images of brutal objectivity, there is the obvious question as to the humanistic possibilities of using such a patently scientific approach. Christian Gosvig Olesen addresses this directly in his Towards a ‘Humanistic Cinemetrics’? essay. While highlighting how cinemetrics can “defamiliarize our objects of study by foregrounding their constructed nature” using methodology that is “primarily scientist” he claims that cinemetric approach does indeed “qualify as humanistic data research but that they still need to be fully recognized and distinguished as such” (39). The goal of the proposed research is to collect and visualize cinematic data (from American horror films over the past century) and analyze it using cultural studies techniques to realize Olesen’s ‘humanistic cinemetrics” so it may be properly recognized for its humanistic/scientific duality common to many digital humanities projects.
While the scholarly research on horror films is vast and has increased in popularity over the past few decades, there is a particular interest in those who survive. In her classic book, Men, Women, and Chainsaws: Gender in the Modern Horror Film (1992), Carol Clover coined the term “Final Girl” for the generally not-so-feminine female who doesn’t quite fit in but who not only survives the murderer’s horrific trail of death but typically vanquishes him (typically, it’s a ‘him’) in the end. The spectators’ interest is intentionally drawn to the narrative of the survivors. This project is designed to complement this “survivor focus” of the Final Girl by drawing scholarly attention to the collateral narrative damage essential for a successful horror films: the too-often quickly forgotten dead.
Finally, it should be noted that there is public and horror fan interest in this proposed project documenting deaths in American horror films. For one, there’s a Lists of Death Wiki that is a good attempt at documenting deaths in horror films. One fan has gone so far as to document a variety of deaths in the top 80 films in the last 40 years using cultural metrics and put the results in a simple graphic. Then there are claims that blacks always die first and that white females are at the highest risk in horror films. With the database created in the proposed project, the data related to these issues of public/fan interest can be rigorously collected and analyzed to affirm or reject amateur hypotheses; i.e., this project will provide not only a cornucopia of data for scholarly cultural studies research, it will also be that rare scholarly project that crosses over and produces a product of intense public/fan interest.
PART 3 (Oct 20): NARRATIVE (Method, Platform, Tools)
A website will be established which will be the central focus of the project. It will include a downloadable phone app, both Android and Apple versions, spectators will use to tag scenes with demographic and other data. These data will be uploaded via the web to an MySQL database on a USF server at the spectator’s convenience. This approach of a web-based database and a user data-gathering application with XML tags is following the well-proven approach of Yuri Tsivian’s Cinemetrics website; however, whereas Tsivian allowed spectators to modify tags, the tags provided will be fixed to ensure meaningful comparisons can be made between films. While this approach may seem specific to horror films, by simply changing the XML tags to document other events of interest, the entire platform can be used to analyze other genres of film or any other video media, such as music videos or political speech.
The general approach will be for the spectator to select a horror film of their choice from a list curated with the guidance of horror film scholars from USF to ensure the most important films are included. To begin with, the top-10 grossing and/or important horror films from each year between 1910 and 2010 will be selected. In consultation with USF film scholars, a list of XML tags will be selected to ensure the proper data is available for scholarly analysis. Once the XML tags and films are selected, the phone application will be developed by computer scientists at USF.
The website will be hosted using Bluehost and created using WordPress. The website will use a MySQL database capability built into Bluehost to establish and collect all the app data, which will be submitted as a file and imported into the database. Note that any issues with Bluehost may be resolved via their excellent, human-based, customer support. Further, Bluehost and WordPress are industry standards with WordPress as the platform for a third of all websites.
For visualization of the data, a WordPress plug-in called “Visualizer” will be used based on its simplicity, proven interoperability, and web-based approach. This will be the initial, rudimentary stage of data visualization for local scholars to use to ensure rapid analysis and subsequent publications – they will not have to wait for the polished version. At a later stage, a custom, user-friendly data visualization approach will be created with paid software developers to write the HTML code for a simpler GUI that better serves a general audience. For mapping data, the initial analysis for scholars will use Google Maps and a later stage may employ a more robust GIS software solution allowing more interaction and expository data for a general audience.
Once scholars have begun to identify specific, significant trends, these data will be documented in static charts and posted on the website. For each of these charts, a scholarly interpretation will be accompany it and a blog, again using the built-in capabilities of Bluehost/WordPress, will allow a general audience to view and respond to the charts. It is expected that these static charts and the data in the database, which may be exported via XML to other databases, will be the core, enduring elements of the project while the phone applications and custom HTML software may decay over time without additional funding. The core of the project may be maintained on Bluehost for a low cost (~$100/yr) once the funding period has expired; however, if no additional funding is forthcoming to update the software, the project is expect to be transferred to USF for archival purposes.
PART 4 (Oct 27): NARRATIVE (Outcomes, Deliverables)
Outcome
The overarching outcome of this project is a deeper understanding of the role death plays in our culture as revealed in the hidden and thus uninvestigated data manifest in a century of American horror films. This project will allow this treasure trove of hidden data to be revealed for the first time. This outcome will be met through the investigation of the multiple technical as well as human metrics selected and graphically presented by film scholars as well as dedicated fans and casual spectators. This project is novel in that anyone may discover an interesting, academically significant relationship, post their results, and begin a discussion where anyone, regardless of academic background, my participate and benefit.
Deliverables
Website – The core deliverable will be a robust, easily maintained, website that allows for integration of data from spectators to populate the database; has the capability to graphically represent data in a manner useful for scholarly investigation as well as easy enough for a casual user to manipulate; has blogging capabilities for user interaction; and includes an artistic element that allows uses to create either digital videos of the “faces” of those who died in the films or a single graphic that includes a subset of the these faces.
Phone Apps – These apps, in both iOS and Android formats, will allow users to input horror film data and upload these data to the website.
Database – The database will capable of being downloaded to any user in an XML format capable of being imported in a user’s database of choice for offline investigation
Source Code – Any compute code created for the project (e.g., graphing routines, data transfer) will be available to all users.
Journal Articles – There will be a single paper created from this project. This paper will document the project itself and the work done to create it as well as document lessons learned for future, similar projects. It will also document a few interesting relationships from the project that are of scholarly interest and provide culturally unique results. This paper will be presented at future Digital Humanities conference in Year 3 of the project.
PART 5 (Nov 3): BUDGET, DATA PLAN I
BUDGET
The total 3-year budget for this project is $380K (maximum grant amount is $530K).
Year 1 – Build Robust Internet Presence: $30K (“Discovery” phase of the grant)
– Subject Matter Expert consultation on American Horror Films: $2.5K
– Subject Matter Expert consultation on Humanities & Cultural Studies: $2.5K
– Website creation with custom Database/graphing integration & testing: $20K
– University Facilities Fees: $5K
Year 2 – Build User Applications & Begin Gathering Data: $100K (“Prototyping” phase of the grant)
– Subject Matter Expert consultation on American Horror Films: $10K
– Subject Matter Expert consultation on Humanities & Cultural Studies: $10K
– Build, certify phone apps & refine website: $75K
– University Facilities Fees: $5K
Year 3 – Refine System, Finish Gathering Data, & Report Results: $250K (“Production” phase of the grant)
– Subject Matter Expert consultation on American Horror Films: $25K
– Subject Matter Expert consultation on Humanities & Cultural Studies: $25K
– Continue building, debugging, testing, certifying app & website: $190K
– Travel to present project paper at 2 DH conferences: $5K
– University Facilities Fees: $5K
DATA PLAN
The data horror film spectators gathers using the phone app answers the following questions:
- Who dies? (demographic data, and a headshot of the deceased taken from the film using the cell phone app)
- Who kills? (human, monster, alien)
- Where? (general geographic location & specific location: indoors [house, business, etc.], outdoors[field, street, etc.])
- How are they killed? (stabbing, mutilation, transformation[a symbolic death])
- When? (time of day/night, time since the start of the film)
All the above data are either numeric, a character string, or a .jpg file (for the headshot). All these data will be input to the database for future data visualization/graphic/mapping queries.
PART 6 (Nov 10): DATA PLAN II
While the intent is to maintain the developed system via a follow-on maintenance grant and/or using university resources, the plan if either of those two do not materialize is to document and distribute that data as follows:
Source Code Data – These are the custom codes for the website’s graphical user interface as well as the custom code that allows the integration of the website with commercial graphics packages and the smartphone applications. The code for the smartphone applications will also be provided in an open source format. All these data will be stored on a free GitHub website and will be freely available for download.
Movie-Specific Data – These are the unique data collected by spectators of the curated horror films on their smartphone applications or input directly into the online database. These data will consist of tagged events at specific timestamps during the film as well as headshots of the characters. The data from this database will be downloaded and posted monthly in an XML format so that it may be imported into the database of the user’s choosing for offline analyses.
Forum Data – These data will be considered the lowest priority for preservation from this project. The movie-specific data is the highest priority for preservation, the source data is second, and the blogging discussions third. While there will certainly be many significant discussions related to the relationships discovered in this project, it is expected that the most significant findings will inform peer-reviewed journal articles. Efforts will be made to archive significant blogs as flat files and store them on the GitHub website for this project.
PART 7 (Nov 17): SUPPORTING MEDIA (Visualizations, Graphs, Maps, Charts, Illustrations)
Since this project relies on the interaction of public to ensure its success as well as the work of humanities scholars and digital media professionals, the grant pursued is the Digital Projects for the Public grant under the National Endowment for the Humanities. This grant is divided into 3 years with funding of $30K for the Discovery year, $100K for the Prototyping year, and $400,000 for the Production year. As stated above, this project will require only $25K for each of its 3 years.
Below is a flowchart for the envisioned flow of data and visualizations in the project.
Amateur horror film fans will enter data using a smarphone app and to capture a headshot from the film of the deceased.
After a user selects a specific variable (e.g., who dies in the films), the location of their death will be plotted on a zoomable map of the United States and hovering over an icon (red dot) will reveal a pop-up with a headshot of the deceased and other information.
PART 8 (Nov 24): SUPPORTING MEDIA + ABSTRACT
ABSTRACT
Brief Abstract
The aim of this project is to carefully document and analyze deaths in the first century of American Horror Films. The public will use a custom phone app to capture demographic and filmic data and download these data to a central website for archiving, visualization, and public/scholar interaction on forums. The goal is to focus on the powerless in American popular culture and uncover the hidden patterns related to their representation over time as well as identify hidden horror filmmaking conventions. The outcome is to develop a deeper historical understanding of the powerless in America and the development of a simple, robust data acquisition system that is easy to sustain and readily available for other media research.
Detailed Abstract
Since the 1910 film Frankenstein, artists have created thousands of American horror films that document human existential fears of dying, either literally or figuratively, at the hands of monsters, aliens, science, humans, or some other phenomena. While much scholarly work exists on specific genres or elements of a subset of these films and objective analyses of quantitative data from films also exists (e.g., cinemetrics), there hasn’t been a project that blends the two. Until now. For this project; digital humanists, technical professionals, and the public will work together to curate and analyze the first century of American horror films (1910-2010) to document the who, what, where, when, and how of human death in these films. Horror film fans will use a custom phone app to gather and submit their data to an online database where scholars and amateurs alike will use visualization tools to create maps, X-Y plots, and histograms. These graphics will reveal, for the first time, the trends and hidden patterns in this horror film archive motivating humanist questions concerning gender, race relations, modernity, and fear as well as exposing changes in horror filmmaking conventions that reflect cultural changes in America. Further, these graphics will be available for public review and discussion on a project forum so horror film fans and scholars may mutually benefit from their interaction. To ensure the core of the project, the database, exists beyond the project’s conclusion, its data will be freely available for download and will be stored on archival websites. The ultimate outcome of this project will be a much deeper understanding of American culture in the 20th century from an overlooked and often trivialized lens that will benefit horror film fans and humanities scholars alike while providing a robust, interactive method for analyzing films of different genres in the future.
PART 9 (Dec 1): PROJECT PITCHES
Below is the PowerPoint presentation used for the final project pitch.