Datasets:

benlehrburger

/

college-text-corpus

like 0

Feature #3: Scale

Architecture can feel uncanny given its relative size to other objects. We consider “familiar” architecture to be that which aligns with our expectations across various features, including size. A sofa chair is quite far from being uncanny, but if it were scaled up to the size of a house, it would violate our expectations and likely elicit an uncanny reaction.

Blade Runner manipulates scale to assert uncanny power dynamics. Figure 6c depicts the Tyrell Corporation’s headquarters from the opening scene of the movie, in which their monolithic pyramid towers over the fiery industrial landscape of post-apocalyptic Los Angeles. The immense scale of the building relative to the rest of the city not only hints to the viewer that the Tyrells are an omnipotent force, but also elicits a visceral, uncanny reaction. The same sense of scale can be felt in Figure 6b, an artificially generated version of the still frame. Using language that insinuates a power dynamic – the pyramid “looms” over the industrial landscape – helps Midjourney scale the monolith relative to its surroundings to generate an uncanny scene.

Citizen Kane also uses scale to insinuate power, which turns the film’s architecture uncanny. The movie’s plot revolves around Charles Foster Kane’s endless struggle to live up to the power he radiates from inside of his castle, Xanadu. Figure 7c shows Orson Welles’ depiction of Xanadu from the movie’s opening scene. It is so massive, and the fog is settled so thick, that it is hard to tell where the castle begins and ends. So too is Figure 7b, the artificially generated version of Xanadu created using Welles’ original description of Xanadu from his screenplay. The descriptive inputs like “fairy-tale,” “early-morning sky,” and “iron grille work” are not innately uncanny, but when the architecture is so big that it becomes uninterpretable, it is easy for it to elicit an uncanny reaction.

One of the most notably uncanny scenes from the movie that we discussed was the “hall of mirrors” scene, in which Charles Kane gets caught in an infinite reflection of himself during his fall from power. Originally, we analyzed the ornate architecture of the mirror in comparison to Kane’s disheveled appearance. Through my analysis, however, I see this scene as uncanny because of its clever use of scale. Figure 8c illustrates the present image. Were Kane just looking at a single reflection, the scene would have significantly less affect because that is the sort of one-dimensional reflection we are used to seeing. However, Kane appears infinitely many times, getting smaller and smaller with each reflection. This effect is nicely artificially generated in Figure 8b as well. Scaling a reflection from one-dimension to infinite dimensions pushes the scene into an unfamiliar uncanniness that I have been searching for in this analysis.

Conclusion

I have used mixed media and artificial text-to-image generation to argue that light, context, and scale are strong elicitors of the uncanny. First, I presented The Jolly Corner, The Third Man, and artificially generated images based on their texts to show how light, darkness, shadows, and silhouettes prompt uncanny experiences. I then turned to The Bridge on the Drina and images created with its text to support the notion that historical context can build uncanny associations with architecture. Lastly, I analyzed scenes in Blade Runner and Citizen Kane to convey how scale builds uncanny power dynamics in architecture and corroborated those findings using Midjourney outputs.

Artificial text-to-image generation is still in its infancy but may come to play a larger role in the architectural realm. It serves as an objective, highly informed source with extremely creative and clear outputs. In this study, I used AI to analyze uncanniness in architecture, but there are a limitless number of other architectural themes that are ripe for analysis. These models have been trained on far more inputs than we could ingest in a lifetime, so it would be a shame to not consider their architectural view. As we’ve witnessed, they may just view architecture in the same way that Ivo Andrić, Orson Welles, Ridley Scott, and William James see the world.

Tuesday evening last week, after learning that I was subject to disciplinary action for my COSC 10 midterms, I drove into Hanover to try to reconnect with Dartmouth. I sat outside of the Shattuck Observatory and surveyed the stars – something I’ve routinely done since taking an astronomy course freshman year. I could see Saturn, Jupiter, and the faint glow of Pluto all in a single field of view above Baker Tower. It was beautiful. Slowly but surely, I began to remember how I got to where I find myself today.

When I began applying to colleges, I knew that I wanted to study somewhere that offered the freedoms I needed to develop into a successful student and independent person. Somewhere that allowed me to experiment with courses and not force me to pursue one area of interest; somewhere different than my parents’ alma maters so I could ensure that getting accepted was a feat of my own doing; somewhere small where interpersonal relationships mattered and individual identity was held in high regard. Given these criteria, Dartmouth was a clear winner. Attracted by its tight-knit community and liberating academic structure, I applied early decision, and was accepted.

I arrived at Dartmouth last fall with enthusiasm, vigor, and slight nerves. I met my roommate Noble Rai on move-in day, and we immediately fell into rhythm. Despite growing up 1500 miles apart – one in Omaha outskirts and one in New York suburbs – we had developed nearly identical traits and interests. We spent many (late-night) hours debating science and religion, cooking carbonara, making naive attempts to reform American government, and rigging our room with levers and pulleys so we could turn off the light switch without getting out of bed. I, a cognitive science major, and Noble, a government and economics enthusiast, supplemented each other’s knowledge and helped the other learn in leaps and bounds. It was precisely the mutualistic friendship I had always craved but never quite attained. So, we decided to live together again sophomore year, and leased a home in Quechee with four of our close friends.

I enrolled in COSC 10 this term to see if I want to minor in computer science. I had been looking for a practical application of cognitive science to pursue professionally, and artificial intelligence surfaced as a great candidate, for which I would need substantial coding experience. Soon after the add/drop period, I learned that Noble had also elected to take COSC 10. I was thrilled. We chose each other as partners on the course’s problem sets, attended virtual lectures from the same computer monitor, shared classnotes, and broke down concepts together post-lecture. When midterms one and two rolled around, we did the vast majority of our studying together and compiled joint exam notes. We reviewed the same material, rewatching lectures and rereading lecture notes. Having lived and studied together over the past year, we were used to bouncing questions back and forth and filling in the gaps of each other’s knowledge. As it would turn out, I let this bond become too close.

Midterm 1 was on Tuesday, October 6. It was a four-part, open-note assessment. Noble and I began studying Thursday, October 1 and reviewed material into late Monday evening. We worked at the dining table, the house’s only desk setup. When we were confident in our understanding, we moved to opposite sides of the table and independently began our exams. We had absolutely no plan to collaborate during the test. I opened Part 1 of the exam around midnight, and according to the Submission Timeline provided by Professor Pierson in the Academic Honor Principle Violation Reporting Form, Noble began a few minutes afterwards. During this multiple choice section, Noble and I did not ask each other any questions about the content of the exam, but did ask each other about where to locate information in our compiled class and exam notes. It is not surprising to me, then, that our answers were the same, considering we were drawing from an identical set of notes. However, I recognize any discussions at all during the exam – no matter the content – are forbidden by the COSC 10 syllabus and Midterm Instructions. I was so used to having open discussions with Noble that I allowed our close relationship to carry over into a time of individual testing. For this, I am guilty.

In between exam sections, Noble and I took breaks to review material that either one of us had felt shaky on in the previous section. Because we took breaks together, and used all forty minutes of allotted time in the subsequent three exam sections, our Submission Timelines are aligned. Parts 2-4 of the exam asked for short answer responses. During these sections, Noble and I conferred about where to find information in our notes, such as which page discusses a specific Data Structure or what lecture covered a particular ADT. Again, we overstepped boundaries in doing so. I do, however, maintain that none of our discussions regarded the exam questions themselves and all code written and submitted on my exam was entirely my own. At no point did Noble or I copy each other’s code or send code to one another. I attribute some of the similarities in Noble’s and my code to drawing from the same set of notes and helping each other find material in those notes. Yet, I attribute the other correlations to conceptual similarities that arose from having worked through nearly all other course content together, which we were allowed to do. Since we had a similar, if not identical understanding (and misunderstanding) of Java concepts prior to the exam, it is not surprising to me that our code turned out analogous on paper.

The events that transpired during Midterm 2, taken on October 27, were similar to those during Midterm 1 but with a few exceptions. Two weeks before the exam, Noble and I were involved in a motor vehicle accident that left me with a concussion and Noble with a fractured neck. After missing classes and needing an extension on Problem Set 3, Noble and I became even more reliant on each other in our studying, which we began the Friday prior to the exam. But when we finished studying and were prepared to take the test – early Monday morning – we assumed our seats at opposite sides of the dining table. Again, Noble and I had no plan to collaborate. I maintain that on Midterm 2, all code written and submitted by me is entirely my own and I did not provide any code to Noble. During Part 1 and Part 2 of the exam, Noble and I again discussed where to find particular information in our notes, but did not discuss any aspect of the exam itself. In between Part 2 and Part 3, Noble and I did not review material together, so our Submission Timeline diverges for Parts 3 and 4. However, on Part 3 of Midterm 2, I inadvertently did not check the box on Question 1 certifying that all work was my own. This was an honest mistake; eager to get to the code itself, I accidentally scrolled right past it. On Parts 3 and 4, I concede no transgressions. Noble and I took these sections at different times, and did not ask each other questions about our notes. The following day, when I realized I hadn’t answered Question 1 of Part 3 after receiving a zero on that section, I sent a follow up email to Professor Pierson verifying my certification.

Noble and I did not share or copy code on Midterm 1 or Midterm 2. We consulted nearly identical exam notes during the test and developed a nearly identical conceptual understanding of course content throughout the term. It makes sense, then, that on Parts 3 and 4 of Midterm 2, in which Noble and I did not confer whatsoever, our code is still strikingly similar. Our note-sharing can account for some similarities in code in other sections, but when we did not confer at all, we still ended up with similar responses. In other questions, it is clear that we did not work in tandem. On some occasions, one of us submitted an answer while the other did not. Had we coded together, I don’t see why or how this would have been the case.

Noble and I did ask each other questions about where to find information in our notes during Midterm 1 on Parts 1-4 and during Midterm 2 on Parts 1-2. At no point did we discuss the exam itself. But, any and all exchanges are clearly forbidden by the COSC 10 Syllabus and the Midterm Instructions. Unfortunately, I allowed my close-natured relationship with Noble to cross the line. Taking the exam at a dining room table across from my best friend felt all too comfortable, and I let my guard down. There are times for proximity and dialogue, and then there are exams. This is something I’ve known since elementary school, and my judgement should have been better. I further detail my methodology during the exam, respond to Professor Pierson’s accusations, and note my transgressions in the attached ‘Annotated Appendix’ document.

I would like to apologize to the Dartmouth faculty, the Committee on Standards, the Computer Science department, its chair, Professor Cormen, my COSC 10 instructor Professor Pierson, my TA Brian Morrison, the other students in my COSC 10 section, and the greater Dartmouth community. I also apologize to my parents, and my younger sister, Emily, who are disappointed in me but know that what happened is not representative of who I am.

I am committed to Dartmouth and the values that it upholds. My goal of attending the College was to learn, explore ideas, make friends, form relationships with faculty, and, yes, have some fun too. But I also sought to become more independent, and by becoming so intertwined with Noble in times where partnership is disallowed, I failed that mission. That said, I have learned an important lesson: I can enjoy a tight-knit relationship, but in class and during exams I am a student first and a friend second. Moving forwards, I will ensure that I am accountable for myself and only myself during testing. I should be more cognizant of exam instructions and never skirt the line of permissibility. While the present issue regards COSC 10, I can say with confidence I will not find myself in any situation like this again.

Spanish: I studied Spanish for quite a number of years (up until college) and always

loved its flow. Its linguistic structure makes it so it can be spoken rapidly, but it still sounds elegant. I incorporated Spanish into my own language for these reasons, and because I wanted to better understand its linguistic basis. I adopted Spanish’s stress for my language. Spanish words are stressed on the last syllable (except in words ending in -/s/, -/n/, or a vowel), and this predictable stress is part of what gives Spanish its pleasing cadence. I wanted to recreate Spanish’s rhythm, so my language has a fixed stressed on the last syllable of a word (I also love the Spanish voiced alveolar trill (/r/) so I used that too) (Spanish, n.d.).

Hawaiian: I heard the Hawaiian language for the first time earlier this year, and was very

taken aback. Even without seeing it written on paper, it was clear from an oral standpoint that the language uses an abundance of vowels. Often times in English, too many consonants complicate the language in my mind, so I wanted to adopt the Hawaiian style of vowel overload. It feels more natural and steadfast. Thus, I took the dominant Hawaiian vowels (a, i) and consonants (/m/, /n/, /l/, /p/, /k/), and also based my phonotactic structure (CVC) off of Hawaiian’s simplistic phonotactic structure (CVV). I hope to reflect the Hawaiian language’s vowel usage in this way (Hawaiian, n.d.).

Arabic: This language is quite foreign to me, but I was always intrigued by it. I have

heard time and time again that Arabic is one of the most beautiful languages in the world, and I wanted to understand why. I learned that Arabic sounds so lovely because of its diversity in sounds. As such, I wanted to incorporate some of this diversity into my own language. So, I adopted the Arabic consonants /θ/, /z/, /t/, and /f/. As for vowels, Arabic has three pairs of corresponding short and long vowels. In my language, I wanted to replicate this diversity by using vowels across a range of heights and depths (Mustafawi, n.d.).

My language has one unique noun class: the hypothetical. A hypothetical noun is one that is not in physical existence, or the speaker is unsure of the noun’s existence. In this way, my language averts the need to distinguish between specific and unspecific nouns using articles. So, if I wanted to talk about the person sitting immediately to my right, I wouldn’t use the hypothetical noun for ‘person.’ But if I wanted to talk about any generic person, I would use the hypothetical noun for ‘person.’

I am from Westchester, New York, where the indigenous population is the Lenape. Lenape is also the name of their language, a member of the eastern Algonquian language family. There are roughly 16,000 Lenape remaining, most concentrated in Oklahoma, Wisconsin, and Ontario after being pushed out of their Northeastern homeland by the expanding European colonies. None are fully fluent speakers of Lenape, but there is a push among younger generations to revitalize the language. In my research, I believe that the word order of Lenape is SVO, but its syntax is not thoroughly studied or well documented (as with most Algonquian languages).

Lenape phonology is quite different from Standard American English. The language consists of 13 consonants, two of which are affricates ([č] and [š]). Lenape vowels are presented into contrasting long-short pairs. In total, there are six pairs (/iː/ /i/, /eː/ /e/, /aː/ /a/, /uː/ /u/ /əː/ /ə/, /ɔː/ /ɔ/).

"Between the house and non-house it is easy to establish all sorts of contradictions. Inside the house, everything may be differentiated and multiplied. The house derives reserves and refinements of intimacy from winter; while in the outside world snow covers all tracks, blurs the road, muffles every sound, conceals all colors. As a result of this universal whiteness, we feel a form of cosmic negation in action. The dreamer of houses knows and senses this, and because of the diminished quality of the outside world, experiences all the qualities of intimacy with increased intensity."

- Bachelard (The Poetics of Space), 40-41

"I have only one dream. It is the oldest of humanity, of man, in time. It is paradise. I would like to give paradise to everyone."

- Frei Otto

"Why should we build very large spaces when they are not necessary? We can design halls spanning several kilometers and covering a whole city, but we have to ask, what does it really make? What does society really need?"

- Frei Otto

French poet Charles Baudelaire speaks on the dichotomy between house and non-house in his 1860 book Les paradis artificiels: while nature dominates the non-house, the house represents a form of "cosmic negation," ignorant of the outside world and capable of its own design and connotation. I will look to use this transformative ability of the house in designing the LISTEN center. The center will function as a "cosmic negation" of visitors' troubling situations in other realms of life; when they visit the center, they should take a momentary mental vacation. Frei Otto -- German architect and Pritzker Prize winner -- will guide my philosophy for designing a resource library for the LISTEN center. Otto stresses material and economic consciousness; given the frequency with which visitor’s go to LISTEN, it is an opportune location to provide them with resources (job board, transportation maps, computers, cookbook library). When their ideas are combined and applied, Baudelaire and Otto will inspire the construction of a comfortable, luxurious, resourceful food pantry.

How would you describe yourself?

My name is Ben Lehrburger. I’m a sophomore at Dartmouth College where I am an intended Cognitive Science major and minor in Quantitative Social Science and Computer Science. I’m compelled to learn why people think what they think and use that understanding to drive insights across industries.

What’s your past market research experience?

Marketing Intern – Patti Conte, Ltd.

Researched media outlets (then pitched artists)

Amassed and analyzed data on competing up-and-coming NY-based PR firms

Found competition does well with local NY artists; learned we were underleveraged because we target east coast outlets but most of our clients are not east coast-based

Project Intern – Herrmann

Assigned AI-driven browser extension project and researching SaaS NLP options

After vetting their capabilities, went with IBM

Why do your skills complement a market research role?

Analytical mind: I always want to answer the “why” and understand the driving factors of a trend; I am an expert at spotting patterns and connecting the dots

Eg. Patti Conte

Data-driven: I find truth in numbers and statistics; I am well-versed with data-analytics platforms and know how to handle that data to derive meaningful insights

Eg. Herrmann current project (statistical analysis of classifier performance)

Expert communicator: I can take a jumble of information and turn it into findings that means something to people; my main area of study is understanding how people understand, so this is my greatest asset

Eg. Sister and 9th-grade teacher using infographic for example

Now learning Tableau

How do your studies prepare you for this position?

Cognitive Science

Learning peoples’ motivating factors, which allows me to make better predictions and better foresee causal relationships between people and markets

Quantitative Social Science

Learning how to take qualitative phenomena and quantify them in a meaningful way

Learning data-analytics with technical tools like R

Computer Science

Gives me particular industry knowledge in a field that has grown in its dominance over the course of my lifetime

What tools do you use?

Excel

R

Learning how to use Tableau

What does a market researcher do?

Study market conditions to examine potential sales of a product or service

What do people want? How much will they pay for it?

[Prep for Marts & Lundy Interview]

Why the nonprofit space?

I want to be in consulting because (1) I like to problem-solve and (2) I want to learn about wide-ranging industries; if I can fulfill those wishes working with commercial or nonprofit organizations, then I would choose to work with nonprofits because I’m doing more problem-solving – ie., helping nonprofits help others

Past experiences: