A Geometric Proof of the Determinant (That No Textbook Gave Me)
Audience:
Overview
This video presents an elementary geometric construction and proof of the determinant formula, developed to fill a gap often left open in introductory and advanced linear algebra literature.
The project originated from a two-year study journey that began in high school after watching 3Blue1Brown’s Essence of Linear Algebra. While the series offers great visual intuition, it relies on the determinant without proving its algebraic formula geometrically. Standard textbooks like Boldrini’s or Axler’s Linear Algebra Done Right either omit the geometric origin or treat the formula as an axiomatic given (e.g., Axler proves equals the product of eigenvalues using polar decomposition , assuming the determinant’s existence and properties beforehand).
Instead of attempting to construct the determinant from eigenvalues, this proof builds the formula ground-up using elementary geometry accessible to high school and early undergraduate students.
Core Mathematical Insights
-
Oriented Height ():
Standard -dimensional volume of a parallelepiped is defined as: This proof constructs the determinant inductively by replacing standard height with oriented height relative to a chosen hyperplane: -
Additivity and Opposite Half-Spaces:
Popular geometric explanations often fail to prove why the additive property (linearity) holds geometrically. This proof demonstrates that for signed volume to be additive across hyperplanes, the measure must yield opposite signs in opposite half-spaces (or half-planoids in ).
Planned 4-Part Series Roadmap
This video is Part 1 of a planned four-part series on -dimensional geometry in :
| Part | Title & Objective |
|---|---|
| Part 1 | Proof of the volume formula for an -dimensional parallelepiped in (This video) |
| Part 2 | Proof of the volume formula for an -simplex in |
| Part 3 | Proof of the volume of an -dimensional parallelepiped in using , where is the Gram matrix |
| Part 4 | Proof of the volume of an -dimensional simplex in |
AI Disclosure Statement
WORKFLOW DISCLOSURE:
- Concept, Proof & Logic: 100% original human work created by the author.
- Storyboard & Visual Design: Designed frame-by-frame in Excalidraw by the author.
- Code Implementation: Claude was used to translate the Excalidraw storyboards and timing specifications into Manim (Python) code.
- Final Rendering: Executed via Manim. I am submitting this full disclosure to the SoME organizers to determine formal competition eligibility regarding AI-assisted coding versus AI-generated visual media.
Analytics
Comments
This explanation is fine. It’s perhaps overexplained at the beginning, which is a shame in the sense that you clearly put a lot of work into the first ~6:00, and I think that work was mostly unnecessary. I think this problem of overdoing the explanations pervades the video, but I’m willing to believe that I only think that because of my background this material (and indeed a great deal of this specific way of motivating the determinant) is quite familiar to me. If you don’t hear something similar from other folks, ignore me. But if you do, then strongly consider paring down your explanations more.
In any case, it’s a good starting place for you. In particular, you seem to have an extremely strong belief that your explanations should be outlined, foreshadowed, and/or signposted, and you are correct. This is the thing that I am constantly telling folks in these reviews to do more of. Your challenge is going to be to let go of worrying you’re not doing enough to pull your audience along, and extend some confidence they will rise to meet you. (WOAH except that pull of the Riesz representation theorem. That’s a heavy bit of machinery to spring on your viewer without warning.)
Finally, I’m willing to believe that the narration is not written by a chatbot, because there are enough idiosyncracies that I’m not used to seeing from, say, Claude-generated text. However, it reads like it was written by someone who spends a lot of time talking to chatbots, because there are obnoxiously many idiosyncrasies which I am used to seeing from Claude-generated text. This is not inherently a problem for the video, but it’s something I wanted to flag because it has the potential to be a problem for you: to be taken seriously in a human-centered field like science communication, you are going to need to develop a expository voice that is credibly human. This takes time, and you’re new at this, so you’re getting there, but you’re not there yet. Keep pushing.