112 Matrix and Multi-Origin High-Dimensional Geometry
598
0
·
2026/04/25
·
4 mins read
☕
WriterShelf™ is a unique multiple pen name blogging and forum platform. Protect relationships and your privacy. Take your writing in new directions. ** Join WriterShelf**
WriterShelf™ is an open writing platform. The views, information and opinions in this article are those of the author.
Article info
This article is part of:
Categories:
⟩
⟩
Date:
Published: 2026/04/25 - Updated: 2026/11/13
Total: 975 words
Like
or Dislike
About the Author
I love science as much as art, logic as deeply as emotion.
I write the softest human stories beneath the hardest sci-fi.
May words bridge us to kindred spirits across the world.
More from this author
More to explore


Matrices and Multi-Origin High-Dimensional Geometry
Abstract: In classical linear algebra, a matrix is ordinarily defined as a two-dimensional array of numbers and serves as the algebraic representation of linear transformations. Within the framework of Multi-Origin High-Dimensional Geometry (MOC), this paper investigates the ontological status of matrices. It demonstrates that matrices are not the ontology of high-dimensional spaces, but projective representations of high-dimensional structures projected onto low-dimensional observation interfaces. From this standpoint, core notions including matrix rank, matrix multiplication, eigenvalues and singular value decomposition admit corresponding geometric interpretations. This framework aims to deepen the understanding of linear-algebraic structures and their role in the representation of high-dimensional spaces.
I. Paradigm Shift: Ontology versus Projection
Conventional linear algebra treats matrices as fundamental mathematical entities. Their tabular two-dimensional form is often identified directly with the structure they describe. While computationally effective when applied to high-dimensional spaces or multivariate systems, their ontological standing is rarely examined.
In the MOC framework, space is regarded as a continuous manifold with multiple origins and intrinsic curvature, which constitutes the underlying ontology. When high-dimensional relations must be expressed on a bounded low-dimensional medium, such as a two-dimensional sheet of paper or a computational grid, a projection operation is required. A matrix is precisely the algebraic manifestation of this projection on the low-dimensional interface.
From this perspective, a matrix is not the high-dimensional space itself, but a representational tool for the high-dimensional ontology on a low-dimensional cross-section.
II. Geometric Origin of the Two-Dimensional Form
Why do matrices typically take the form of two-dimensional tables? In the classical setting, this follows from definition; within MOC, it corresponds to the requirement for minimal complete representation of relational networks on a low-dimensional interface.
Consider the relation between two origins A and B. To fully describe their mutual interaction, such as dominance or coupling strength, at least two components are required: the action of A upon B and the action of B upon A. This naturally forms a 2\times 2 array.
Generalising to n origins, mutual relations over all ordered origin pairs are fully encoded by n\times n entries. Accordingly, the two-dimensional array is no accident: it is the minimal complete topological structure required for relational networks under low-dimensional projection.
III. Geometric Correspondence of Matrix Multiplication
In classical theory, matrix multiplication is defined as the composition of linear transformations. Under MOC, this operation receives a geometric interpretation: matrix multiplication represents the successive superposition of two low-dimensional projections.
Suppose high-dimensional ontology A projects to the low-dimensional matrix M_1, and high-dimensional ontology B projects to the low-dimensional matrix M_2. The product M_1\times M_2 observed by an observer corresponds to successive geometric action of the two projected results.
The non-commutativity of matrix multiplication (M_1 M_2 \neq M_2 M_1) reflects the geometric non-commutativity of projection sequences: different ordering of operations yields distinct final low-dimensional outcomes.
IV. Reinterpretation of Core Concepts
From the “matrix-as-low-dimensional-projection” viewpoint, central matrix concepts may be interpreted as follows:
1. Rank
- Classical definition: The maximum number of linearly independent rows or columns.
- MOC interpretation: The maximal dimension of the high-dimensional ontology that the low-dimensional projection can recover without information loss. Full rank signifies a complete projection permitting inversion back to the original space; rank deficiency indicates folding of the high-dimensional structure during projection, where part of the information is discarded.
2. Eigenvalues and Eigenvectors
- Classical definition: Vectors retaining their direction under the transformation, together with their scaling factors.
- MOC interpretation: Eigenvalues correspond to stationary points of equilibrium configurations of the high-dimensional ontology under low-dimensional projection. The eigenvalue spectrum, the set of all eigenvalues, is an invariant under projective transformation and encodes intrinsic properties of the ontology for a given viewing perspective.
3. Determinant
- Classical definition: Volume scaling factor induced by the linear transformation.
- MOC interpretation: The compression ratio when the high-dimensional ontology is projected onto a low-dimensional interface. A zero determinant means the projection direction lies orthogonal to the effective dimension of the ontology, causing complete collapse of information on that cross-section.
4. Singular Value Decomposition (SVD)
- Classical definition: Factorisation of an arbitrary matrix into a product of rotation, scaling and rotation.
- MOC interpretation: Any low-dimensional projection decomposes into observer orientation, intrinsic ontology structure, and secondary observer orientation. Singular values quantify the fraction of information preserved by the projection along each principal axis.
V. Explanation of Three Fundamental Questions
This framework offers unified answers to several foundational questions in linear algebra:
1. Why can matrices represent high-dimensional data?
Because a matrix is itself a projection. Writing high-dimensional data into matrix form amounts to selecting a low-dimensional observation interface and performing projection; matrix entries record the landing positions of projection rays.
2. Why is matrix multiplication defined in this particular way?
The row-times-column rule is the only self-consistent algebraic form for successive superposition of projections. It is not an arbitrary human convention, but a necessary consequence of geometric action in low dimensions.
3. Why may a single high-dimensional structure correspond to multiple distinct matrices?
Because the viewing angle differs. The same multi-origin space yields different matrices when projected onto different low-dimensional interfaces, yet these matrices share invariant features: identical eigenvalue spectra (intrinsic invariants), equal rank (projection completeness), and identical singular values (information retention ratios). This supplies the geometric foundation for the invariance of eigenvalues under similarity transformations.
VI. Conclusion
The MOC framework furnishes a new lens for understanding matrices: matrices are not ultimate entities of reality, but projected slices left by high-dimensional ontology on low-dimensional interfaces.
This interpretation does not diminish the value of matrices. On the contrary, their indispensable role across mathematics and engineering arises precisely from their efficiency in carrying projected information of high-dimensional structures.
From this perspective, the operational rules and properties of matrices can all be seen as natural mappings of high-dimensional geometric structures onto low-dimensional cross-sections. This provides a valuable geometric complement for ontological enquiries into linear algebra.