Ever since the modern philosophy represented by Descartes’ argument – I think, therefore I am – began, the central philosophical issue has been the question of what human beings can know and how it is possible. Noteworthy here is that epistemic sovereignty belongs to the thinking subject “I.” In other words, an epistemological foundation results from intuitive self-consciousness, which cannot be denied its existence and can serve as an indubitable fundamental truth for our knowledge. This kind of Descartes’ philosophical position, such as foundationalism, subjectivism, and intuitionism, has influenced the formation of the Cartesian anxiety. And I think this anxiety could arise from DHers’ craving for a theoretical ground for Digital Humanities.  

After reading “Developing Things: Notes toward an Epistemology of Building in the Digital Humanities,” I want to ask: Do DHers need their epistemic foundation for their DH works? Suppose we could define DH as an activity of building and making something with humanistic inquiry. In that case, digital works, such as computing, modeling, and visualizing, should be enough to be regarded as scholarship. If we place epistemic sovereignty on the community of inquirers (We) instead of I, what matters would be whether we could secure and maintain an open structure of our inquiry process. Digital tools are tools. They don’t have to be a theory. (For example, the coding language Python is merely a tool. Should it be a theory?)

The problem DHers face is not an epistemic controversy but a power game between a newly developing field with new methods and tools and those existing humanities fields. The theory-practice dichotomy shouldn’t be a problem here, as digital tools provide new humanistic perspectives. Instead, DHers can expand their academic area beyond the scope of traditional humanities with digital methodologies. Within the community of inquirers, DHers don’t have to be truth-seekers and shouldn’t put the Cartesian anxiety on their shoulders; becoming “a kibitzer or a therapist or an intellectual historian” (Rorty 1982) should be the first step for DHers to take toward digital scholarship and making new vocabularies of their own -building, modeling, and computing- should be one of their central jobs.

This week’s readings opened my eyes to the complexities of collecting and visualizing data, and helped me to understand that data I’ve taken to be neutral is anything but. Many of the questions raised for me in Johanna Drucker’s “Humanities Approaches to Graphical Display” were answered in the rest of the readings, but I want to linger on some of these questions and their solutions in my post. 

Drucker introduces the idea that capta should be interpreted by a factor of “x”, where is x can be the point of view of the data collector, agendas, presumptions, assumptions, or other conventions that lend a subjective dimension to the capta. This made me wonder if advanced digital technologies can capture the complex, interpretive qualities of data by letting us view datasets with different factors of “x” highlighted through animations. If qualitative data could be displayed in a way that brings attention to the variable priorities of the data collectors, viewers, and content, then maybe our general conception of data could be loosened from the rigid idea that data is self-evident and even approach the understanding that our positionality influences how we interpret the statistical world around us. 

The complexities and necessary considerations of data visualization gave me pause: if graphical representations of interpretive data are so insufficient when it comes to representing nuance, then why do we even need them? Jennifer Guiliano and Carolyn Heitman’s “Difficult Heritage and the Complexities of Indigenous Data” helped me to understand that recognizing the messiness of data is essential in turning a critical eye towards data. When describing ethical modes of accessing information about Indigenous communities, the authors state that “there is no expedient or universal solution” (15). This is a good thing, because it means that those holding and accessing information are forced to consider the impact of their data use on Indigenous communities that have a right to safeguard information. Guiliano and Heitman’s assertion applies to the complexities of data visualization raised by Drucker, too: the fact that it’s difficult to display context makes both the creator and the viewer of a data visualization think deeply about the assumptions that underlie a more straightforward graph. By embracing the fact that there is no universal or expedient way to display capta, we can draw attention to the shortcomings of non-humanist means of collecting, analyzing, and displaying data and interrogate what’s at stake – what is sacrificed – in the pursuit of expediency.

I have a lot to say about the Difficult Heritage piece out of this week’s readings on Data and Visualization so I’m just going to focus on that one. 

I’m taking another class on archival research practices at the moment and just the other day I was telling my professor about how making knowledge publicly accessible comes with it’s own set of issues and I didn’t have the words to explain it until I read this paper. I had started to notice how public access makes it so that an individual can bypass having to have ties to a specific community in order to access the information and knowledge of the community that has been preserved. This article really highlights how Indigenous communities are susceptible to experiencing harm when researchers outside of their communities gain access to their data, culture, and knowledges. 

I also really appreciated the point about how, “a more systemic approach to the traditional knowledge labels employed by Mukurtu [allows] for individual items and collections to be withheld from view not just to members of the public but also to members of the originating community who might not be of the appropriate clan, stature, gender, or position.” To me this point also applies to Indigenous scholars since even those of us within Indigenous communities need to be careful not to overstep by taking on projects that are not our place to take on due to our positionality within the community. I also found the discussion about “the heritage effect” / “the museum effect” to be very useful to the development of one of my current research projects since I will be talking about how this effect shows up in the Mexican-American community (e.g. people knowing they have indigenous ancestry yet distance themselves from the current reality indigenous peoples in Mexico and speak of indigenous people as if they only exist in the past).

Additionally, the discussion in this article about how, “a researcher might need to navigate permissions for use of the data by the institutions holding the material, family or clan members with an interest in the materials, the tribal cultural heritage officer charged with preserving the tribe’s history, as well as the tribe’s governing authority,” made me think about the upcoming film Killers of the Flower Moon (directed by Martin Scorsese starring Leonardo DiCaprio and Lily Gladstone). The film tells a story that takes place on the Osage Nation’s territory and the director actually (allegedly) did a great job of working with the Osage community to get the film to be something that many members of this indigenous community approve of. The director (in my opinion based on what I know about the film-making process) exemplified an approach that Indigenous scholars have been asking of non-Indigenous people to show when doing research or any kind of the work that requires gaining access to and telling Indigenous stories (I recommend checking out this Twitter thread by the former chief of the Osage nation sharing his experience with the film-making process if you’re interested, you’ll probably have to log in to Twitter to access the full thread though).

I started this mapping praxis to show the correlation between gentrification and the reduction in the laundromat business. However, I soon realized that the project was too ambitious for an outsider who still sees data as metaphysical entities and understands digital tools as a magic wand. This post is a failed journey, but I hope this will help those with a similar background.

Step #1_Hypothesis

Hypothesis 1: Each form of doing laundry is income-related, i.e., social class, in NYC.

1. Upper Class+ : It doesn’t bother these people because someone else will care for them.
2. Upper Class– : This group has built-in washers and dryers in their newly built condo units. It means they don’t have to go outside to do laundry.
3. Middle Class: They have to go outside, but fortunately, they don’t have to go outside the building. They have laundry facilities in the basement of the building.
4.  Lower Class+ : Using a small laundromat located outside of the building. People in this group might live in the same town as group 3, but usually rent a unit in an old family house.
5. Lower Class- : Using a Mega-Laundromat (30+ washers & 30+ dryers) operated by the quarters  (sometimes by laundry card).

 I wanted to focus on group 5, which has relatively more vulnerable housing conditions and likely have more new development or construction project in the neighborhood. 

Hypothesis 2: The people living in gentrified areas in NYC could have different problems in their day-to-day lives besides rent increases. Gentrification might be forcing (especially  “Mega-“) laundromats out of business.

Step 2_Choosing a Mapping Tool

After reading <Finding the Right Tools for Mapping>, I chose to use Carto for the mapping praxis because the reading introduces Carto as an “intuitive and relatively easy to use” tool. However, it turns out that this introduction does not apply to me:

• I don’t know the technical terms they use, such as CSV, KML, GeoJSON. I finally figured out that these are certain forms of data files.
• Using Carto as expected requires essential skills – setting boundaries & sorting data out – to handle data sets first. (That I don’t have.)
  • My old laptop’s capability is not good enough to smoothly operate this tool on the web.

Step #3_Data Access    

To verify my hypothesis, I tried to find two data sets: Income-related Housing data & Laundromat business data.

I investigated two websites: EquityNYC – Eviction Filing database & OpenData NYC – Legally Operating Business database from DCA(Department of Consumer Affairs).

• I downloaded the Excel file for <Eviction Filing>, but the <Laundromat Business database> link does not work. (OpenDataNYC does not seem to maintain the link between the website and DCA’s database.)
  • I have a dataset with eviction filing by community district, but I couldn’t find the boundary file to map this dataset.

Step #4_Side Project: My Laundry Life in NYC

Instead, I decided to make my NYC laundry life map:

• I made Google Sheets with the information on laundromats’ addresses.
• I got the latitude and longitude information for each address from Google Maps.
• I uploaded the CSV file to Carto and got the visualized map.

https://clausa.app.carto.com/map/3a5cd6e2-2fd9-467e-9e11-e3c24a5b8adb

At the beginning of using Carto for mapping praxis, I thought I could automatically upload some datasets and get a visualized map. I didn’t know I had to upload an Excel (CSV) file to have the first layer of the map where I could mark and decorate. After many days of struggling, I finally learned how to make a mappable Excel file.

While reading Jennifer Guiliano and Carolyn Heitman’s article, “Difficult Heritage and the Complexities of Indigenous Data,” I was struck by how the open access movement — an ostensibly well-intended shift toward a more democratic approach to using/sharing data — can pose a threat to historically marginalized communities by perpetuating colonial practices of producing and preserving knowledge.

In Catherine D’Ignazio and Lauren Klein’s book Data Feminism, the authors describe a map that I believe exemplifies what Guiliano and Heitman frame as an “Indigenous-centric approach” to (re)constructing cultural memory. The map was researched and designed by Dr. Margaret Pearce, in collaboration with First Nations, Métis, and Inuit communities across Canada, and is titled “Coming Home to Indigenous Place Names in Canada.” The boundaries of the mapped landmass mirror that of the Anglo-Western conception of Canada, but the place names on the map reflect those used historically and contemporaneously by the Indigenous communities that inhabit(ed) those lands.

At multiple points, the map protects or obscures knowledge, as a means of preserving the privilege of knowledge for a select few, rather than as a right for all. The exact locations of place names are not given, so as to limit the ability of outsiders to gain access. (This is achieved through the map’s massive scale of 1:5,000,000, which allows for a certain level of descriptive ambiguity.) Furthermore, in some cases, place names are not translated into English. 

Additionally, as a measure to prevent decontextualization and misuse, the map may not be reproduced at all unless the relevant communities are consulted. As Klein and D’Ignazio describe in Data Feminism, “Each time the map is reproduced…Pearce writes to the communities to whom the names belong, explains the proposed context of the names, and requests permission for the names to be reproduced in that context.” This allows for and maintains the agency and oversight of the communities who have consented to the use of their place names — and disrupts the tendency for open-access platforms to allow “for objects to be divorced from their conditions of production and contexts of interpretation for all forms of reuse” (Guiliano 18).

When I first learned about Bitcoin about seven years ago, I fantasized about quitting my job, going coding school and embarking on a world changing future Blockchain project for a well-funded tech start-up that not only had stock options but paid salaries with crypto.  Before taking the giant leap of quitting my job, I went to an introductory meeting for a coding boot camp, of which there were many at that time, for an idea of what was in store for me.  This first step was a big letdown, because I left the session feeling that I didn’t have a good knowledge base, the academic prerequisites or skills required to take the class.  In order to go to bootcamp, I would need to take Python and learn other coding skills prior to being admitted, which I wasn’t willing to do at the time. Needless to say, I never quit my job or pursued Bitcoin mining, but this didn’t stop me from doing a little investing….

Fast forward to this intro to DH class where we are exposed to digital methodologies and are required to “learn some of the fundamental skills used often in digital humanities projects,” including skill workshops, as the one I recently attended called: Introduction to Python with Rebecca Krisel. The class was much less intimidating than expect and the instructor was very deliberate in emphasizing that like with any foreign language, one needs to practice, make mistakes and with time and effort, muscle memory will kicks in.

I’ve installed Python on my PC. I’ll start by learning terms and replicating some of what Rebecca went over in the class. I’ve also downloaded a children’s book from the Mina Rees Library for reference. I look forward to the follow-up Python class(es0, but I have started playing around with some easy exercises and here’s what I’ve come up with thus far for days 1 and 2 of practice:

I started off with a simple enough idea: to programmatically map locations in a text.

I went to a store in London and then went to Lake Erie to go fishing.

The idea was to extract “London” from the text and then place it on a map. This proved to be more complex than I anticipated. Both SpaCy and NLTK offer entity recognition through pre-built models that do a decent job of extracting the location (in this case, London) from the text but aren’t equipped to add context.

After running through several possibilities, I came across Mordecai, a project authored by MSU professor Andy Halterman. It uses a SpaCy model to extract locations, it then runs the text through another model that predicts the context based on the other words in the text. In the case of our example, London would be correctly classified as London, the city in Ontario, Canada because of the contextual clue “Lake Erie.”

{   'doc_text': 'I went to a store in London and then went to Lake Erie to '
                'go fishing.',
    'event_location_raw': '',
    'geolocated_ents': [   {   'adm1_count': 1.0,
                               'admin1_code': '08',
                               'admin1_name': 'Ontario',
                               'admin1_parent_match': 0,
                               'admin2_code': '',
                               'admin2_name': '',
                               'alt_name_length': 3.2188758248682006,
                               'ascii_dist': 0.0,
                               'avg_dist': 0.11450381679389314,
                               'city_id': '6058560',
                               'city_name': 'London',
                               'country_code3': 'CAN',
                               'country_code_parent_match': 0,
                               'country_count': 1.0,
                               'end_char': 29,
                               'feature_class': 'P',
                               'feature_code': 'PPL',
                               'geonameid': '6058560',
                               'lat': 42.98339,
                               'lon': -81.23304,
                               'max_dist': 0.2692307692307692,
                               'min_dist': 0.0,
                               'name': 'London',
                               'score': 0.9999490976333618,
                               'search_name': 'London',
                               'start_char': 23},

I also tried with the following variations (there’s a Thames River close to both Londons):

A. I went to a store in London and then went to Thames River to go fishing.

B. I went to a store in London and then went to River Thames to go fishing.

The first example, returned two results London, England and London, Ontario with London England scoring slightly higher. The second example only returned London, England.

Given more time, I would have liked to try mapping the Black Mountain corpus. For this assignment, I chose to work with just one poem: “I, Maximus of Gloucester, to You” by Charles Olson.

After extracting the location data with Mordecai, I saved it as a CSV file, and I used Folium (an open-source mapping library) to map the datapoints

import pandas as pd
import folium

# read in the poem
df = pd.read_csv('poem.csv')

# map data to folium
m = folium.Map(location=[-6.1753924, 106.8271528], zoom_start=2)

# add markers to map
for i in range(0, len(df)):
    folium.Marker(
        location=[df.iloc[i]['lat'], df.iloc[i]['lon']],
        popup=df.iloc[i]['text'],
        icon=folium.Icon(color='green')
    ).add_to(m)

Mordecai extracted:

• Oregon, Ohio
• New England, Jamaica
• Gloucester, England

While context was added to each of the locations found in the poem, only one of them was correct. I attempted a similar process with four other poems with similar results.

While this tool is useful when the cues in the text are obvious (or more likely common, e.g. Paris and the Eiffel Tower) the Mordecai model would need to be tuned differently for this type of text.

Some interesting questions that were raised in the process of completing this assignment were:

• Could it be possible to extract location based on dialectical differences?
• Mordecai was funded by the CIA as part of the Political Instability Task Force. What are the ethical implications of using a tool like this for alternative purposes?
• Creative processes like poetry seem to be at odds with the statistical methods used in Natural Language Processing or machine learning. What kind of patterns are worth looking for?
• In poetry, geographic references serve a variety of purposes. A map, while appearing multi-dimensional, can flatten literary analysis to just coordinates. Could a map reveal something new?

After comparing the tools described in “Finding the Right Tools for Mapping,” I opted to try mapping with Carto because its interface seemed relatively intuitive, and I was able to get started with a free two-week trial. I’m also looking forward to experimenting with QGIS in the future and appreciated learning how transferable its interface is to ArcGIS (which I decided to hold off on experimenting with because of the associated costs).

While getting familiar with Carto, I reflected on our readings and felt particularly inspired by Mayukh’s “Dividing Lines” and Brown’s “Slave Revolt in Jamaica, 1760-1761: A Cartographic Narrative.” Both of these pieces/projects highlight the ways in which maps (and their creators) selectively place value on places and spaces, and they can both enhance and limit our understanding. A map is not neutral and technology is not foolproof, as Mayukh details:

Brierley’s return to home may well have been impossible without Google Earth. But “Lion” represents this process as a one-way transaction between an error-prone, sleepless human and an intelligent device, rather than as a human’s struggle to overcome a potentially useful technology’s limitations and biases. In the process, the film flattens the inherently complicated relationship we all have to such platforms, the ways they do not meet our expectations, the ways they occasionally disappoint us, the ways they are inflected by our socioeconomic status and the assumptions and stereotypes that can guide corporate strategies.

https://reallifemag.com/dividing-lines/

This reminded me of some of my own frustrations over the years with Google Maps. When I used Google Maps in San Francisco, I realized quickly that it didn’t note topography. What might be a flat, easy ten minute walk in NYC can actually be a strenuous climb in a city of hills. Similarly, in parts of Morocco and Barbados, Google Maps often failed to accurately represent the walkable pathways commonly used throughout their cities, rendering it almost useless for navigation. This prompted my thinking of the other ways Google Maps may be limiting my experience of space and navigation. For example, Google Maps currently lets me label places like “home” and “work,” and I can save locations in lists, but I have little flexibility beyond that to mark places of personal importance. I became inspired to try.

I decided to attempt to map “something that is not necessarily (or traditionally thought of as) mappable” and was curious as to whether I could map something like the human emotions I might experience on one of my regular walking routes. I realized that in my daily walks to and from the train, I might experience a number of fleeting emotions, prompted by the space I was traversing. For example, I might leave home content and then get annoyed with pedestrian traffic at a certain intersection or feel nostalgic while passing by a certain location that sparks a fond memory. You can check out my attempt here.

Carto made it relatively easy to upload an excel document with latitudes and longitudes associated with categorized emotional experiences. With more time and experience in Carto, I’d be interested in being more diligent about exploring how my emotions respond to mappable places and/or building a heat map of the sentimental value I (and others) attach to places as a complimentary piece of information to attach to existing and future listings.

I was a geography major in college and “I’m old enough to” have studied cartography before “personal computers and electronic publishing,” to quote Mark Monmonier. As I recall, the analog map that I created ‘back in the day,’ was called The Houses of Worship in Delaware County, Ohio, where I attended college, and it was created using tools that included a compass, cartography paper, planimeters, stencils, pencils, erasers and dividers.

Despite my college major, I am neither a cartographer nor a geographer, but this latest assignment has caused me to examine how I use maps or non-traditional maps on a day-to-day basis outside of travel. I came up with one prime example and that’s in my current career in finance: namely, in the transportation sector of fundamental equity research and specifically in the subsectors of trucking and logistics.

Since Covid, the trucking and logistics parts of the economy have been of great concern for research analysts as well as everyday retail consumers. According to Ernst & Young, the “COVID-19 pandemic accelerated preexisting issues in the supply chain and brought priorities such as visibility, resilience and digitization to the fore.” What we closely observe in fundamental research are data points that affect the forward earnings of stocks that we follow and those for which we publish equity (stock) research. Some of the data inputs that were of particular concern during COVID were bottlenecks in ocean freight scheduling, port congestion, worker shortages, pricing constraints and warehousing capacity. Since earlier this year, as the logistics of moving products has eased, the shift in interest has been around inventory stocking and the health of the consumer. These and other current economic issues can at times be mapped, but given the many variables needed, we often source data from third-party vendors to arrive at month-over-month trends, which we can then use to provide our investors with projections (albeit some may be “distorted”) on where we value stocks. In this current economic cycle, trucking (including truck brokers) is facing a difficult path into 2024. Many trucking companies are cutting costs in a high diesel fuel cost environment (a large contributor to shrinking margins) as they are experiencing lower demand from retailers, their main clients, because they already may be stuck with seasonal inventory that is taking up warehouse space. Furthermore, the UAW (and previous labor negotiations) have been and are additional risks to the future earnings for labor and products dependent sectors…in this instance autos and drivers, respectively. The pandemic inventory stocking went from a “just-in-case” over stocking environment to a “just-in-time” for seasonal products state – one that was customary pre-pandemic. How would one map these data points and concerns?

Well, as I mentioned earlier, my firm uses third party vendors for certain data and in my attempt to recreate a map, I’ve chosen one particular data point and that’s tender volume. The map below reflects tender volume, which is a load offer from a shipper to a trucking company. The national trucking companies can use this heat map to determine where they may find the highest volumes to keep their truckers busy and those areas where less staff is needed. The higher volume areas also allow the trucking companies to charge more for deliveries. It is noteworthy that most of the map indicates steady or declines in tender volumes – perhaps a good indication of a slowing economy – something that economists watch, but also a consideration for fundamental research analysts.

I tried to replicate this map with my free version of Tableau, but I’m still not skilled in re-creating the data inputs.