On the Clifford Algebraic Description of the Geometry of a 3D Euclidean Space

Jayme Vaz Jr.; Stephen Mann

REVIEW 3 major objections 4 minor 19 references

On the Clifford Algebraic Description of the Geometry of a 3D Euclidean Space

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper shows that the Clifford algebra Cℓ3,3 expresses every standard 3D geometric transformation — reflection, rotation, translation, shear, non-uniform scale, and perspective projection — from a single paravector-based model.

desk verdict A solid Cℓ3,3 reformulation of the authors' paravector model with a real, fixable sign error in the headline perspective theorem — worth refereeing, not worth trusting as printed. read the letter →

arxiv 1908.08110 v1 pith:YIQVTDHF submitted 2019-08-21 math.GM

Jayme Vaz Jr. , Stephen Mann This is my paper · ORCID

classification math.GM MSC 15A66

keywords CliffordalgebraCl(33)paravectorperspectiveprojectioncotranslationHodgestargeometrictransformationprojective

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper is trying to establish that the Clifford algebra $\mathcal{C}\ell_{3,3}$ of the quadratic space $\mathbb{R}^{3,3}$ is a single algebraic setting in which the everyday transformations of three-dimensional Euclidean geometry — reflection, circular and hyperbolic rotation, translation, shear, non-uniform scale, perspective projection, and pseudo-perspective — all appear as the same kind of operation. The route is to embed each vector $\vec v$ as the half-sum $v=\frac{1}{2}(v_+ + v_-)$ of a positive-signature and a negative-signature vector, and to represent a point by a paravector $P=1+p$. From that embedding, each transformation is a versor sandwich $\Phi P \tilde \Phi$, except perspective, which uses a newly defined Hodge-star operation, cotranslation, composed with translation. If this is right, the smaller algebras $\mathcal{C}\ell_{3,0}$ and $\mathcal{C}\ell_{0,3}$ remain sufficient for reflection and rotation, but the full $\mathcal{C}\ell_{3,3}$ is needed for the rest, and the whole graphics pipeline can be discussed in one Clifford-algebraic language.

What carries the argument

The central object is the real Clifford algebra $\mathcal{C}\ell_{3,3}$, generated by three vectors $e^+_i$ with positive square and three vectors $e^-_i$ with negative square, all mutually anticommuting. Inside it, the paper makes two load-bearing constructions: the embedding $\gamma(\vec e_i)=\frac{1}{2}(e^+_i+e^-_i)$ that turns ordinary vectors into halved sums of a positive and a negative part, and the paravector representation $P=1+p$ of a point. The mechanism that carries the arguments is the sandwich action $P\mapsto \Phi P\tilde \Phi$ for invertible $\Phi$, which produces reflection, rotation, hyperbolic rotation, shear, non-uniform scale, and translation, plus the Hodge-dual sandwich $P\mapsto \star^{-1}[T(\star P)\tilde T]$ called cotranslation, which produces perspective and pseudo-perspective. The Hodge star is defined by $\star A_k = \langle \widetilde{A_k}\,\Omega_{\mathbb{V}}\rangle_{3-k}$ with $\Omega_{\mathbb{V}}=e_1e_2e_3$. Theorems 1 through 9 supply the explicit versors, and Theorem 14's infinitesimal classification — only $k=0$, $k=1$, and $k=2$ with $\psi_2=a\wedge b^*$ — is what lets the paper identify the resulting general transformation as projective.

What would settle it

Implement, in coordinates, the paper's perspective formula $(T_{\vec e}\circ W_{\vec n/a}\circ T_{-\vec e})(P-E)$ for a point $P$ in front of the eye and a non-axis-aligned normal $\vec n$, and verify that the resulting weighted point lies on the plane $\vec x\cdot\vec n=c$ with the predicted weight; a failure for any such input would show that the cotranslation/Hodge-star mechanism is not producing genuine perspective projection.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that $\mathcal{C}\ell_{3,3}$ is not just another Clifford model of 3D space but the one in which the natural identification $v=\frac{1}{2}(v_+ + v_-)$ makes affine and projective point transformations look alike. Points are paravectors $P=1+p$, and the operations are explicit versors: $N=n_+ n_-$ for reflection, $R=e^{\theta(u_+ v_+ - u_- v_-)/2}$ for circular rotation, $H=e^{\eta(u_- v_+ + v_- u_+)/2}$ for hyperbolic rotation, $S=e^{t(u_+ + u_-)(v_+ - v_-)/4}$ for shear, $D=e^{t u_- u_+/2}$ for non-uniform scale, and $T=e^{v/2}$ for translation. For perspective, the paper defines cotranslation $W_{\vec v}(P)=\star^{-1}[e^{v/2}(\star P)e^{v/2}]$ using the Hodge star built from the trivector $\Omega_{\mathbb{V}}=e_1e_2e_3$, and proves that perspective projection from an eye point to a plane is $(T_{\vec e}\circ W_{\vec n/a}\circ T_{-\vec e})(P-E)$, with pseudo-perspective given by $W_{\vec n}$. Reflection and rotation act separately on the $+$ and $-$ parts and therefore reduce to $\mathcal{C}\ell_{3,0}$ or $\mathcal{C}\ell_{0,3}$; all other operations mix the two parts, which is why the full algebra is needed. The paper also classifies the infinitesimal transformations, showing that only scalar, vector, and simple bivector $a\wedge b^*$ generators preserve the paravector condition, and concludes that the formalism realizes all 3D projective transformations.

Load-bearing premise

The load-bearing premise is that ordinary 3D vectors really are represented by the half-sum $v=\frac{1}{2}(v_+ + v_-)$ inside $\mathcal{C}\ell_{3,3}$ and that points are faithfully represented by paravectors $P=1+p$; if this identification does not capture the affine and projective structure of points, every transformation theorem in the paper loses its geometric meaning.

Editorial extensions

If this is right

Because $N$ and $R$ preserve the $+$ and $-$ sectors, reflection and rotation can still be implemented in $\mathcal{C}\ell_{3,0}$ or $\mathcal{C}\ell_{0,3}$, exactly as in the classical formulations.
Hyperbolic rotation, shear, non-uniform scale, translation, and cotranslation mix the two sectors, so these operations cannot be reduced to the smaller algebras and require the full $\mathcal{C}\ell_{3,3}$.
In a graphics pipeline, an affine part $A$ followed by a perspective part $T$ acts on a point as $P'=\star^{-1}[T(\star(AP\tilde A))\tilde T]$, meaning the affine and perspective stages are two versors rather than one.
Because $v^2=0$ for the embedded vector, the translation versor $e^{v/2}$ gives a square-root-of-point factorization $1+p=e^{p/2}1e^{p/2}$.
Translation is the one operation that cannot be rewritten as a cotranslation-style single versor; the paper shows any attempt generates unwanted quadratic terms.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Beyond the paper: the square-root factorization suggests a natural exponential interpolation between points in $\mathcal{C}\ell_{3,3}$, a use the paper does not develop.
Beyond the paper: the $+/-$ splitting has the same algebraic shape as particle/hole splitting in a two-level fermion system, so operator normal-ordering methods could automate the composition identities of Section 5.
Beyond the paper: Theorem 14's infinitesimal classification is directly testable — generate the finite group from scalar, vector, and simple-bivector generators and compare it with the full projective group; any projective transformation that is missed identifies a place where the direct versor channel is insufficient and cotranslation must be used.
Beyond the paper: since the model deliberately works in six signed dimensions rather than eight, a concrete extension is to add one null pair of dimensions and check whether quadric-surface transformations of the larger model can be recovered without losing the $\mathcal{C}\ell_{3,3}$ core.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Bottom line: this is a careful, mostly checkable reformulation of the authors' earlier paravector model in Cℓ3,3, and the advertised perspective theorem is wrong as printed. The stress-test calculation is correct: with a = c + n·e, the point produced by Theorem 8 lies on x·n = c + 2e·n, not on the perspective plane. The sign should be a = c − n·e. So the abstract's headline application is currently unsupported, though the fix appears small and local.

What the paper does well: the Cℓ3,3 embedding γ(v) = ½(v+ + v−), the Hodge-dual cotranslation W_v(P) = ⋆⁻¹[e^{v/2}(⋆P)e^{v/2}], and the infinitesimal classification in Theorems 13–14 are the real content. Theorems 1–7 are derived in the text, and the reflection, rotation, hyperbolic rotation, shear, scale, translation, and cotranslation computations are explicit and coherent. The comparison with Goldman–Mann's R(4,4) is honest and useful, and the claim that this lives in Cℓ3,3 rather than R(4,4) is legitimate.

Soft spots. Theorem 8 and Theorem 9 are both pointer proofs ("similar to [9]"), which is a normal practice but becomes load-bearing here because Theorem 8 is false as stated. A corrected theorem with an actual proof needs to appear. The reliance on [9] for the paravector identification is substantial but not circular: the embedding is defined and used in this paper, and the transformation theorems are worked out on its own terms. There is also a factor-of-two slip in the translation proof's intermediate formulas that cancels in the final result; that is minor. Section 6's matrix interpretation is a plausibility argument for the infinitesimal generators rather than a full classification, but the appendix does support the classification it claims.

Who should read it: anyone comparing ambient Clifford algebras for affine and projective transformations, especially in geometric algebra for computer graphics. The Cℓ3,3 catalogue is useful; Theorem 8 should not be cited as a proof of perspective projection until corrected.

Recommendation: this deserves peer review. A serious referee should require a corrected and proved Theorem 8, a proof or exact reference for Theorem 9, and a cleanup of the translation calculation; the rest can stand.

Referee Report

3 major / 4 minor

Summary. The paper develops a Clifford-algebra model of 3D Euclidean point geometry based on Cℓ3,3. Points are represented as paravectors P=1+p, with vectors embedded through v=1/2(v+ + v-). The paper derives versor expressions for reflection, circular and hyperbolic rotation, shear, non-uniform scale, and translation, and introduces a Hodge-dual based cotranslation operation W_v(P)=⋆^{-1}[e^{v/2}(⋆P)e^{v/2}]=P+g(p,v). It claims that perspective projection is a composition of translation and cotranslation (Theorem 8) and that pseudo-perspective is obtained by cotranslation (Theorem 9). Section 6 analyzes infinitesimal transformations and argues that the framework yields all projective transformations, and the paper compares the model with the R(4,4) framework of Goldman and Mann.

Significance. If the central claims held, this would be a useful addition to geometric-algebra tools for computer graphics: it gives explicit, parameter-free algebraic formulas for a wide range of affine and projective transformations in a single algebra, and the matrix interpretation in Section 6 makes contact with standard homogeneous-coordinate practice. The derivations for Theorems 1–7 are largely carried out explicitly and are checkable. However, the advertised perspective application (Theorem 8) is false as stated because of a sign error in the definition of a, and Theorems 8 and 9 are stated without proofs. The paper therefore needs substantive revision before the claims can be accepted.

major comments (3)

[5.7.1, Theorem 8] The stated formula is not a perspective projection. For q=p-e and d=q·n, the composition (T_e∘W_{n/a}∘T_{-e})(P-E) gives the paravector d/a + q + (d/a)e, whose location is X=e+(a/d)q. Hence X·n=e·n+a. With a=c+n·e this is c+2n·e rather than c, so X lies on the wrong plane whenever n·e≠0; for example, n=e3, c=1, E=(0,0,2), P=(0,0,0), the formula yields X=(0,0,5) instead of (0,0,1). The correct parameter is a=c-n·e. Since the proof is omitted and this is the advertised central application, the theorem must be corrected and proved.
[5.7.2, Theorem 9] Theorem 9 is stated without proof and deferred to [9], and it invokes a “point at infinity” that is not defined in the paravector model of this paper. The pseudo-perspective claim in the abstract therefore rests on an unverified and undefined assertion. A proof, or at least a precise definition of points at infinity in this Cℓ3,3 model, plus a verification that W_n(E) produces such a point, should be included.
[3 and 4] The identification γ(v)=1/2(v+ + v-) and the paravector representation P=1+p are load-bearing assumptions of the paper, and the text itself says the justification for the identification comes from Section 4 of [9]. Because all transformation theorems in Section 5 are computed inside this identification, the manuscript should either give a self-contained derivation of the natural map or explicitly state this as a standing assumption taken from [9].

minor comments (4)

[5.7.1] The same symbol P is used both for a point and for the perspective plane in the statement of Theorem 8, which makes the statement harder to read.
[Appendix A] The heading “Proof of Theorem 6.2” should refer to the numbered theorem in Section 6, which is Theorem 14 in the present numbering.
[5.8] The last paragraph contains a doubled word: “translation and and cotranslation”.
[3.1] In the definition of δ^{σ,σ'}_{[ij,ab]}, the factor δ_{ia}δ_{ib} appears to be a typo for δ_{ia}δ_{jb}; the surrounding computation only works with the second reading.

Circularity Check

0 steps flagged · score 2.0 of 10

No formal circularity: the Cℓ3,3 transformation theorems are derived from stated embedding; self-citation and omitted perspective proofs are support gaps, not circularity.

full rationale

The core derivation chain is not circular: the reflection, rotation, hyperbolic rotation, shear, non-uniform scale, translation, and cotranslation theorems (Theorems 1–7) are proved in the text from the stated embedding γ(v) = 1/2(v+ + v−) and the paravector representation P = 1 + p; none of these results is obtained by fitting, by renaming an input, or by defining the conclusion into the premise. There are no fitted parameters and no empirical predictions that reduce to their inputs. The paper does, however, lean on the authors' prior paper [9] in several places: Section 5.7.1 omits the proof of the perspective projection theorem ("The proof is similar to the one in [9], and we omit it here"), Section 5.7.2 defers the pseudo-perspective theorem to [9], and Section 7 says that the key vector identification map is justified by the natural map of [9]. These are self-citation burdens and omitted-proof gaps, especially for the advertised perspective application, but they are not equation-level circularity: the formulas are not defined in terms of the conclusions they are meant to establish, and the central theorems are derived independently in the text. Separately, Theorem 8 as printed is a correctness risk: with a = c + n·e, the computed location lands on x·n = c + 2n·e rather than on x·n = c, so the perspective formula appears to need a sign correction. That is a mathematical correctness issue, not a circularity issue. Overall, no step reduces by construction, and the self-citation burden warrants a low non-zero score.

Assumptions & free parameters 0 free parameters · 5 assumptions · 1 invented entities

No numerical parameters are fitted. The load-bearing structure is a set of algebraic identifications and conventions, plus a dependency on the authors' prior paper for two proofs.

assumptions (5)

standard math Cℓ3,3 is generated by e+_i and e-_i with e+_i e+_j + e+_j e+_i = 2δ_ij, e-_i e-_j + e-_j e-_i = -2δ_ij, and e+_i e-_j + e-_j e+_i = 0.
Standard Clifford algebra background invoked in Section 3.
domain assumption Points are represented as paravectors P = 1 + p, with p the vector from the origin and the scalar part interpreted as weight.
Adopted from [9] in Section 4; the whole transformation formalism is built on this point model.
domain assumption The vector space V3 is embedded in Cℓ3,3 by γ(v) = v = 1/2(v+ + v-), and V3* by v* = 1/2(v+ - v-).
Section 3 states this identification underpins the model; it depends on a chosen basis and Euclidean metric.
standard math The Hodge star is defined with respect to Ω_V = e1e2e3 using the normalization in eq. (3.4), with ⋆^{-1} = ⋆ for n = 3.
Standard Hodge duality restricted to the V3 part; the factor of 2 is a convention chosen to simplify contractions.
domain assumption Theorems 8 and 9 are accepted from [9] without proof in this paper.
Sections 5.7.1 and 5.7.2 say the proofs are similar to [9] and omit them; the central perspective projection application therefore depends on the prior paper.

invented entities (1)

Cotranslation W_v(P) = ⋆^{-1}[e^{v/2}(⋆P)e^{v/2}]
purpose: A dual-space translation operation used to construct perspective projection and pseudo-perspective from ordinary translation.
Defined in this paper, building on [9], and used to reproduce the known perspective projection formula; it has no external falsifiable prediction beyond the internal consistency of the construction.

how reviews work

0 comments

Cite this review

Pith. "Pith review of On the Clifford Algebraic Description of the Geometry of a 3D Euclidean Space." pith.science (2026). https://pith.science/paper/YIQVTDHF

@misc{pith2026190808110,
  author       = {Pith},
  title        = {Pith review of: On the Clifford Algebraic Description of the Geometry of a 3D Euclidean Space},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YIQVTDHF}},
  note         = {Machine review of arXiv:1908.08110}
}

abstract

We discuss how transformations in a three dimensional euclidean space can be described in terms of the Clifford algebra $\mathcal{C}\ell_{3,3}$ of the quadratic space $\mathbb{R}^{3,3}$. We show that this algebra describes in a unified way the operations of reflection, rotations (circular and hyperbolic), translation, shear and non-uniform scale. Moreover, using the concept of Hodge duality, we define an operation called cotranslation, and show that the operation of perspective projection can be written in this Clifford algebra as a composition of the translation and cotranslation operations. We also show that the operation of pseudo-perspective can be implemented using the cotranslation operation. An important point is that the expression for the operations of reflection and rotation in $\mathcal{C}\ell_{3,3}$ preserve the subspaces that can be associated with the algebras $\mathcal{C}\ell_{3,0}$ and $\mathcal{C}\ell_{0,3}$, so that reflection and rotation can be expressed in terms of $\mathcal{C}\ell_{3,0}$ or $\mathcal{C}\ell_{0,3}$, as well-known. However, all other operations mix those subspaces in such a way that they need to be expressed in terms of the full Clifford algebra $\mathcal{C}\ell_{3,3}$. An essential aspect of our formulation is the representation of points in terms of objects called paravectors. Paravectors have been used previously to represents points in terms of an algebra closely related to the Clifford algebra $\mathcal{C}\ell_{3,3}$. We compare these different approaches.

Figures

Figures reproduced from arXiv: 1908.08110 by the authors.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages

[9]

Vaz, Jr., J., Mann, S.: Paravectors and the Geometry of 3D Euclid ean Space. Adv. Appl. Cliﬀord Algebras 28: 99 (2018)

work page 2018
[1]

SIGGR APH Comput

Shoemake, K.: Animating rotation with quaternion curves. SIGGR APH Comput. Graph. 19, 245–254 (1985)

work page 1985
[2]

S ynthesis Lec- tures on Computer Graphics and Animation, Morgan & Claypool Publis hers (2010)

Goldman, R.: Rethinking Quaternions - Theory and Computation. S ynthesis Lec- tures on Computer Graphics and Animation, Morgan & Claypool Publis hers (2010). 30

work page 2010
[3]

Morgan-Kaufmann (2007)

Dorst, L., Fontijne, D., Mann, S.: Geometric Algebra for Compute r Science. Morgan-Kaufmann (2007)

work page 2007
[4]

IEEE Computer Graphics and Applications 20, 76–84 (2000)

Goldman, R.: The ambient spaces of computer graphics and geome tric modeling. IEEE Computer Graphics and Applications 20, 76–84 (2000)

work page 2000
[5]

ACM Trans

Goldman, R.: On the algebraic and geometric foundations of compu ter graphics. ACM Trans. Graph. 21, 52–86 (2002)

work page 2002
[6]

Goldman, R., Mann, S.: R(4, 4) as a computational framework for 3-dimensional computer graphics. Adv. Appl. Cliﬀord Algebras 25, 113–149 (2015)

work page 2015
[7]

Dorst, L.: 3D oriented projective geometry through versors o f R3. Adv. Appl. Clif- ford Algebras, 4, 1137–1172 (2016)

work page 2016

Show all 19 references

[8]

Du, J., Goldman, R., Mann, S.: Modeling 3D Geometry in the Cliﬀord Alge bra R(4, 4). Adv. Appl. Cliﬀord Algebras 27, 3029–2062 (2017)

2017
[10]

Oxford University Press (2016)

Vaz Jr., J., Rocha Jr., R.: An Introduction to Cliﬀord Algebras and Spinors. Oxford University Press (2016)

2016
[11]

Cambridge Universit y Press, 2nd ed

Lounesto, P.: Cliﬀord Algebras and Spinors. Cambridge Universit y Press, 2nd ed. (2001)

2001
[12]

Cambr idge University Press (1995)

Porteous, I.: Cliﬀord Algebras and the Classical Groups. Cambr idge University Press (1995)

1995
[13]

Academic Press, Sa n Diego (1990)

Reese Harvey, F.: Spinors and Calibrations. Academic Press, Sa n Diego (1990)

1990
[14]

M.: Geometric equivalence of Cliﬀord algebras

Botman, D. M.: Geometric equivalence of Cliﬀord algebras. J. Mat h. Phys. 47, 123504 (2006)

2006
[15]

A.: Barycentric Calculus in Euclidean and Hyperbolic Geo metric

Ungar, A. A.: Barycentric Calculus in Euclidean and Hyperbolic Geo metric. World Scientiﬁca Publ. Co., Singapore (2010)

2010
[16]

Chapman & Hall/CRC Computer Graphics, Geometric Modeling , and Animation Series, CRC Press (2009)

Goldman, R.: An Integrated Introduction to Computer Graphic s and Geometric Modeling. Chapman & Hall/CRC Computer Graphics, Geometric Modeling , and Animation Series, CRC Press (2009)

2009
[17]

Vaz, Jr., J.: On Paravectors and their Associated Algebras. Ad v. Appl. Cliﬀord Algebras 29: 32 (2019)

2019
[18]

Klawitter, D.: A Cliﬀord algebraic approach to line geometry. Adv. Appl. Cliﬀord Algebra 24, 713736 (2014)

2014
[19]

In Dorst, L

Gunn, C.: On the homogeneous model of Euclidean geometry. In Dorst, L. Lasenby, J. (eds.), Guide to Geometric Algebra in Practice, pages 297–328, S pringer, (2011). 31

2011

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [9]

Vaz, Jr., J., Mann, S.: Paravectors and the Geometry of 3D Euclid ean Space. Adv. Appl. Cliﬀord Algebras 28: 99 (2018)

work page 2018

[2] [1]

SIGGR APH Comput

Shoemake, K.: Animating rotation with quaternion curves. SIGGR APH Comput. Graph. 19, 245–254 (1985)

work page 1985

[3] [2]

S ynthesis Lec- tures on Computer Graphics and Animation, Morgan & Claypool Publis hers (2010)

Goldman, R.: Rethinking Quaternions - Theory and Computation. S ynthesis Lec- tures on Computer Graphics and Animation, Morgan & Claypool Publis hers (2010). 30

work page 2010

[4] [3]

Morgan-Kaufmann (2007)

Dorst, L., Fontijne, D., Mann, S.: Geometric Algebra for Compute r Science. Morgan-Kaufmann (2007)

work page 2007

[5] [4]

IEEE Computer Graphics and Applications 20, 76–84 (2000)

Goldman, R.: The ambient spaces of computer graphics and geome tric modeling. IEEE Computer Graphics and Applications 20, 76–84 (2000)

work page 2000

[6] [5]

ACM Trans

Goldman, R.: On the algebraic and geometric foundations of compu ter graphics. ACM Trans. Graph. 21, 52–86 (2002)

work page 2002

[7] [6]

Goldman, R., Mann, S.: R(4, 4) as a computational framework for 3-dimensional computer graphics. Adv. Appl. Cliﬀord Algebras 25, 113–149 (2015)

work page 2015

[8] [7]

Dorst, L.: 3D oriented projective geometry through versors o f R3. Adv. Appl. Clif- ford Algebras, 4, 1137–1172 (2016)

work page 2016