Home Colored multizeta values in positive characteristic
Article
Licensed
Unlicensed Requires Authentication

Colored multizeta values in positive characteristic

  • Ryotaro Harada EMAIL logo
Published/Copyright: April 5, 2024

Abstract

In 2004, Thakur introduced a positive characteristic analogue of multizeta values. Later, in 2017, he mentioned the two colored variants which are positive characteristic analogues of colored multizeta values in his survey of multizeta values in positive characteristic. In this paper, we study one of those two variants. We establish their fundamental properties, that include their non-vanishing, sum-shuffle relations, 𝑡-motivic interpretation and linear independence. For the linear independence results, we prove that there are no nontrivial k ̄ -linear relations among the colored multizeta values with different weights.

MSC 2020: 11M38; 11J72; 11J93

Award Identifier / Grant number: JP22J00006

Funding statement: The author also thanks the JSPS Research Fellowships, JSPS Overseas Research Fellowships and the National Center for Theoretical Sciences in Hsinchu for their support during the research project. This work is also supported by JSPS KAKENHI Grant Number JP22J00006.

A Appendix

In this section, we state the detailed proof of Theorem 5.7.

A.1 Key lemma

Lemma A.1

Let V l be a finite k ̄ -linearly dependent subset of CZ w . Then V l is a finite k â€Č -linearly dependent subset of CZ w .

Proof

We set V l = { Z 1 , 
 , Z m } . Without loss of generality, we can assume that m ≄ 2 by Theorem 3.1 and

(A.1) dim k ̄ Span k ̄ ⁥ { V l } = m − 1 .

Again, we can assume that Z 1 ∈ Span k ̄ ⁥ { Z 2 , 
 , Z m } ; then, by assumption (A.1), { Z 2 , 
 , Z m } is a linearly independent set over k ̄ . By Proposition 5.5, we take the matrix Ί j ∈ Mat d j ⁥ ( k ̄ ⁹ [ t ] ) and the column vector ψ j ∈ Mat d j × 1 ⁥ ( E ) ( 1 ≀ j ≀ m ) so that the 4-tuple ( Ί j , ψ j , Z j , r j ) satisfies Lemma 5.4 (i)–(iv).

We define the direct sums of any matrices M 1 , 
 , M m and any column vectors v 1 , 
 , v m whose entries belong to C ∞ ⁱ ( ( t ) ) by

⚁ i = 1 m M i : = ( M 1 M 2 ⋱ M m ) and ⚁ i = 1 m v i : = ( v 1 tr , 
 , v m tr ) tr ,

respectively.

By using this notation, we define the block diagonal matrix Ω and column vector 𝜓 as follows:

Ί : = ⚁ j = 1 m Ί j and ψ : = ⚁ j = 1 m ψ j .

By Lemma 5.4, we have

(A.2) ψ | t = Ξ = ⚁ j = 1 m ( 1 π ̃ w , 
 , a j ⁹ b j ⁹ Z j π ̃ w ) tr

for some a j ∈ F ̄ q × , b j ∈ k × and

(A.3) ψ | t = Ξ q N = ⚁ j = 1 m ( 0 , 
 , 0 , a j ⁹ c j N ⁹ ( b j ⁹ Z j π ̃ w ) q N ) tr

for some c j ∈ F q × , N ∈ N . Using Theorem 5.3, there exist row vectors f j = ( f j ⁹ 1 , 
 , f j ⁹ d j ) ∈ Mat 1 × d j ⁹ ( k ̄ ⁹ [ t ] ) ( j = 1 , 
 , m ) so that if we put F = ( f 1 , 
 , f m ) , then we have

F ⁹ ψ = 0 , f 1 ⁹ d 1 | t = Ξ = 1 and f j ⁹ i | t = Ξ = 0 for ⁹ 1 ≀ i < d j ,  1 ≀ j ≀ m .

Since f 1 , d 1 | t = Ξ is the coefficient of a 1 ⁹ b 1 ⁹ Z 1 / ( π ̃ w ) in ( F ⁹ ψ ) | t = Ξ = 0 and by assumption (A.1), Z 1 is expressed by nontrivial k ̄ -linear combinations of Z 2 , 
 , Z m . We write F â€Č : = ( 1 / f 1 ⁹ d 1 ) F and d : = ∑ j = 1 m d j . Note that the vector F â€Č is of the form

F â€Č = ( f 11 â€Č , 
 , f 1 ⁹ d 1 â€Č , 
 , f m ⁹ 1 â€Č , 
 , f m ⁹ d m â€Č ) ∈ Mat 1 × d ⁹ ( k ̄ ⁹ ( t ) ) ,

where f 1 ⁹ d 1 â€Č = 1 . We have the following by F ⁹ ψ = 0 and f j ⁹ i | t = Ξ = 0 :

(A.4) F â€Č ⁹ ψ = 0 and f j ⁹ i â€Č | t = Ξ = 0 for ⁹ 1 ≀ i < d j ,  1 ≀ j ≀ m .

By using Lemma 5.4, for r = lcm ⁥ ( r 1 , 
 , r m ) , we obtain F â€Č ⁣ ( − r ) ⁹ Ί ⁹ ψ = ( F â€Č ⁹ ψ ) ( − r ) = 0 and thus

(A.5) F â€Č ⁹ ψ − F â€Č ⁣ ( − r ) ⁹ Ί ⁹ ψ = ( F â€Č − F â€Č ⁣ ( − r ) ⁹ Ί ) ⁹ ψ = 0 .

The last column of the matrix Ί j is ( 0 , 
 , 0 , 1 ) tr for each 𝑗, and consequently, the d 1 -th entry of row vector F â€Č − F â€Č ⁣ ( − r ) ⁹ Ί is zero since the d 1 -th entries of the row vectors F â€Č and F â€Č ⁣ ( − r ) ⁹ Ί are 1. The ∑ i = 1 j d i -th column of Ί is

Then the ∑ i = 1 j d i -th entry of the row vector F â€Č ⁣ ( − r ) ⁹ Ί is f j ⁹ d j â€Č ⁣ ( − r ) . Thus, the ∑ i = 1 j d i -th entry of F â€Č − F â€Č ⁣ ( − r ) ⁹ Ί is written as follows:

f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) for ⁹ j = 1 , 
 , m .

We claim that f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) = 0 for j = 2 , 
 , m . Indeed, if there exist some 2 ≀ j ≀ m such that f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) ≠ 0 , we can derive the contradiction in the following way.

Let us take a sufficiently large N ∈ N so that ( f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) ) | t = Ξ q N ≠ 0 and all entries of ( F â€Č − F â€Č ⁣ ( − r ) ⁹ Ί ) are regular at t = Ξ q N . By using (A.3) and substituting t = Ξ q N in (A.5), we obtain

{ ( F â€Č − F â€Č ⁣ ( − r ) ⁹ Ί ) ⁹ ψ } | t = Ξ q N = { ( F â€Č − F â€Č ⁣ ( − r ) ⁹ Ί ) } | t = Ξ q N ⁹ ⚁ j = 1 m ( 0 , 
 , 0 , a j ⁹ c j N ⁹ ( b j ⁹ Z j π ̃ w ) q N ) tr = ∑ j = 1 m ( f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) ) | t = Ξ q N ⁹ a j ⁹ c j N ⁹ ( b j ⁹ Z j π ̃ w ) q N = 1 π ̃ w ⁹ ∑ j = 1 m ( f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) ) | t = Ξ q N ⁹ a j ⁹ c j N ⁹ ( b j ⁹ Z j ) q N = 0 .

Thus, combined with f 1 ⁹ d 1 â€Č − f 1 ⁹ d 1 â€Č ⁣ ( − r ) = 1 − 1 = 0 , we obtain the nontrivial k ̄ -linear relations among Z 2 q N , 
 , Z m q N as follows:

∑ j = 2 m ( f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) ) | t = Ξ q N ⁹ a j ⁹ c j N ⁹ ( b j ⁹ Z j ) q N = 0 .

Then, by taking the q N -th root of the relation, we obtain the following nontrivial k ̄ -linear relation among Z 2 , 
 , Z m :

∑ j = 2 m { ( f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) ) | t = Ξ q N ⁹ a j ⁹ c j N } 1 q N ⁹ b j ⁹ Z j = 0 .

This contradicts our assumption that { Z 2 , 
 , Z m } is a k ̄ -linearly independent set.

Therefore, we obtain f j ⁹ d j â€Č − f j ⁹ d j â€Č ⁣ ( − r ) = 0 for j = 2 , 
 , m , and this equation shows the following:

(A.6) f j ⁹ d j â€Č ∈ F q r ⁹ ( t ) ( j = 2 , 
 , m ) .

By substituting t = Ξ in equation F â€Č ⁹ ψ = 0 , equations (A.2) and (A.4) enable us to obtain the following equalities:

( F â€Č ⁹ ψ ) | t = Ξ = ( 0 , 
 , 0 , f 1 ⁹ d 1 â€Č | t = Ξ , 
 , 0 , 
 , 0 , f m ⁹ d m â€Č | t = Ξ ) ⁹ ⚁ j = 1 m ( 1 π ̃ w , 
 , a j ⁹ b j ⁹ Z j π ̃ w ) tr = ∑ j = 1 m ( f j ⁹ d j â€Č | t = Ξ ) ⁹ a j ⁹ b j ⁹ Z j π ̃ w = 1 π ̃ w ⁹ ∑ j = 1 m ( f j ⁹ d j â€Č | t = Ξ ) ⁹ a j ⁹ b j ⁹ Z j = 0 .

By f 1 ⁹ d 1 â€Č = 1 and (A.6), we have the following nontrivial k â€Č -linear relation among Z 1 , 
 , Z m :

∑ j = 1 m ( f j ⁹ d j â€Č | t = Ξ ) ⁹ a j ⁹ b j ⁹ Z j = 0 .

Therefore, we obtain the desired claim. ∎

A.2 Proof of Theorem 5.7

Proof

We can assume that w l > ⋯ > w 1 without loss of generality. By the definition, CZ w i is a finite set for each i = 1 , 
 , l , and thus, its subset V i is also finite. Let each V i consist of { Z i ⁱ 1 , 
 , Z i ⁱ m i } , where Z i ⁱ j ∈ CZ w i ( j = 1 , 
 , m i ) are the same total weight w i . The proof is by induction on 𝑙.

We require on the contrary that { 1 } âˆȘ ⋃ i = 1 l V i is a k ̄ -linearly dependent set, and then we proceed to our proof by assuming the existence of nontrivial k ̄ -linear relations involving V l .

For 1 ≀ i ≀ l and 1 ≀ j ≀ m l , by combining Theorem 4.4 with Proposition 5.5, we can take r i ⁹ j > 0 , the matrix and the column vector

Ί i ⁹ j ∈ Mat d i ⁹ j ⁹ ( k ̄ ⁹ [ t ] ) and ψ i ⁹ j ∈ Mat d i ⁹ j × 1 ⁹ ( E )

so that d i ⁹ j ≄ 2 and that each ( Ί i ⁹ j , ψ i ⁹ j , Z i ⁹ j , r i ⁹ j ) satisfies Lemma 5.4.

For Ί i ⁹ j and ψ i ⁹ j ( 1 ≀ i ≀ l , 1 ≀ j ≀ m l ), we define the following block diagonal matrix and column vector:

Ί ̃ : = ⚁ i = 1 l ( ⚁ j = 1 m i ( t − Ξ ) w l − w i Ί i ⁹ j ) and ψ ̃ : = ⚁ i = 1 l ( ⚁ j = 1 m i Ω w l − w i ψ i ⁹ j ) .

Here we used ⹁ in the same manner of the definitions of Ω and 𝜓 in the proof of Lemma A.1.

By the requirement in the beginning of proof, { 1 } âˆȘ ⋃ i = 1 l V i is a linearly dependent set over k ̄ . Thus, there exists a nonzero vector ρ = ( v 11 , 
 , v 1 ⁹ m 1 , 
 , v l ⁹ 1 , 
 , v l ⁹ m l ) such that

ρ ⋅ ( ψ ̃ | t = Ξ ) = ρ ⋅ ⚁ i = 1 l ⚁ j = 1 m i ( 1 π ̃ w l , 
 , a i ⁹ j ⁹ b i ⁹ j ⁹ Z i ⁹ j π ̃ w l ) tr = 1 π ̃ w l ⁹ ( v 11 , 
 , v 1 ⁹ m 1 , 
 , v l ⁹ 1 , 
 , v l ⁹ m l ) ⁹ ⚁ i = 1 l ⚁ j = 1 m i ( 1 , 
 , a i ⁹ j ⁹ b i ⁹ j ⁹ Z i ⁹ j ) tr = 0 ,

where v i ⁹ j ∈ Mat 1 × d i ⁹ j ⁹ ( k ̄ ) for 1 ≀ i ≀ l and 1 ≀ j ≀ m i . Then we have the following nontrivial k ̄ -linear relation:

( v 11 , 
 , v 1 ⁱ m 1 , 
 , v l ⁱ 1 , 
 , v l ⁱ m l ) ⁱ ⹁ i = 1 l ⹁ j = 1 m i ( 1 , 
 , a i ⁱ j ⁱ b i ⁱ j ⁱ Z i ⁱ j ) tr = 0 .

In the beginning of this proof, we assumed that there exist nontrivial k ̄ -linear relations between V l and { 1 } âˆȘ ⋃ i = 1 l − 1 V i , and thus, for some 1 ≀ s ≀ m l , the last entry of v l ⁹ s is nonzero. By using Theorem 5.3, we have F : = ( f 11 , 
 , f 1 ⁹ m 1 , 
 , f l ⁹ 1 , 
 , f l ⁹ m l ) , where f i ⁹ j = ( f i ⁹ 1 , 
 , f i ⁹ d i ⁹ j ) ∈ Mat 1 × d i ⁹ j ⁹ ( k ̄ ⁹ [ t ] ) for 1 ≀ i ≀ l , 1 ≀ j ≀ m i and it satisfies F ⁹ ψ ̃ = 0 and F | t = Ξ = ρ . The last entry of f l ⁹ s is a nontrivial polynomial because the last entry of v l ⁹ s is not zero. We choose a sufficiently large N ∈ Z so that f l ⁹ s | t = Ξ q N ≠ 0 and Ο q N = Ο for all factors ζ A ⁹ ( s ; Ο ) which appear in all Z i ⁹ j ( 1 ≀ i ≀ l , 1 ≀ j ≀ m i ). We rewrite the equation ( F ⁹ ψ ̃ ) | t = Ξ q N = 0 as follows by using Ω | t = Ξ q N = 0 , Lemma 5.4 (iv) and the definition of ψ ̃ :

( F ⁹ ψ ̃ ) | t = Ξ q N = ( f 11 , 
 , f 1 ⁹ m 1 , 
 , f l ⁹ 1 , 
 , f l ⁹ m l ) | t = Ξ q N ⁹ ⚁ i = 1 l ⚁ j = 1 m i Ω w l − w i | t = Ξ q N ⁹ ( 0 , 
 , 0 , a i ⁹ j ⁹ c i ⁹ j N ⁹ ( b i ⁹ j ⁹ Z i ⁹ j π ̃ w i ) q N ) tr

when i ≠ l , Ω w l − w i | t = Ξ q N = 0 and then

= ( f l ⁹ 1 , f l ⁹ 2 , 
 , f l ⁹ m l ) | t = Ξ q N ⁹ ⚁ j = 1 m l ( 0 , 
 , 0 , a l ⁹ j ⁹ c l ⁹ j N ⁹ ( b l ⁹ j ⁹ Z l ⁹ j π ̃ w l ) q N ) tr = ∑ j = 1 m l ( f l ⁹ d l ⁹ j | t = Ξ q N ) ⁹ a l ⁹ j ⁹ c l ⁹ j N ⁹ ( b l ⁹ j ⁹ Z l ⁹ j π ̃ w l ) q N = 0 .

Thus, we obtain the following nontrivial k ̄ -linear relation with some f l ⁹ d l ⁹ s ≠ 0 :

∑ j = 1 m l ( f l ⁱ d l ⁱ j | t = ξ q N ) ⁱ a l ⁱ j ⁱ c l ⁱ j N ⁱ ( b l ⁱ j ⁱ Z l ⁱ j ) q N = 0 .

Therefore, by taking the q N -th root of the above relation, we obtain the following nontrivial k ̄ -linear relation for { Z l ⁹ 1 , 
 , Z l ⁹ m l } :

∑ j = 1 m l { ( f l ⁱ d l ⁱ j | t = ξ q N ) ⁱ a l ⁱ j ⁱ c l ⁱ j N } 1 q N ⁱ b l ⁱ j ⁱ Z l ⁱ j = 0 .

This shows that V l is a k ̄ -linearly dependent set. Then, from Lemma A.1, it follows that V l is a k â€Č -linearly dependent subset. However, this contradicts the condition that V l is the k â€Č -linearly independent set. Therefore, Theorem 5.7 holds. ∎

Acknowledgements

The author gratefully acknowledges Professor Dinesh Thakur for informing him of the definition of CMZVs, Professor Chieh-Yu Chang for providing him fruitful comments about the project, and the anonymous referee for making suggestions and comments that have greatly improved the former version of this paper.

  1. Communicated by: Freydoon Shahidi

References

[1] G. W. Anderson, W. D. Brownawell and M. A. Papanikolas, Determination of the algebraic relations among special Γ-values in positive characteristic, Ann. of Math. (2) 160 (2004), no. 1, 237–313. 10.4007/annals.2004.160.237Search in Google Scholar

[2] G. W. Anderson and D. S. Thakur, Tensor powers of the Carlitz module and zeta values, Ann. of Math. (2) 132 (1990), no. 1, 159–191. 10.2307/1971503Search in Google Scholar

[3] G. W. Anderson and D. S. Thakur, Multizeta values for F q ⁱ [ t ] , their period interpretation, and relations between them, Int. Math. Res. Not. IMRN 2009 (2009), no. 11, 2038–2055. 10.1093/imrp/rnp010Search in Google Scholar

[4] T. Arakawa and M. Kaneko, On multiple 𝐿-values, J. Math. Soc. Japan 56 (2004), no. 4, 967–991. 10.2969/jmsj/1190905444Search in Google Scholar

[5] C.-Y. Chang, A note on a refined version of Anderson–Brownawell–Papanikolas criterion, J. Number Theory 129 (2009), no. 3, 729–738. 10.1016/j.jnt.2008.10.012Search in Google Scholar

[6] C.-Y. Chang, Linear independence of monomials of multizeta values in positive characteristic, Compos. Math. 150 (2014), no. 11, 1789–1808. 10.1112/S0010437X1400743XSearch in Google Scholar

[7] C.-Y. Chang, Frobenius difference equations and difference Galois groups, 𝑡-motives: Hodge Structures, Transcendence and Other Motivic Aspects, EMS Ser. Congr. Rep., European Mathematical Society, Berlin (2020), 261–295. 10.4171/198-1/4Search in Google Scholar

[8] C.-Y. Chang, Y.-T. Chen and Y. Mishiba, On Thakur’s basis conjecture for multiple zeta values in positivecharacteristic, Forum Math. Pi 11 (2023), Paper No. e26. 10.1017/fmp.2023.26Search in Google Scholar

[9] C.-Y. Chang and Y. Mishiba, On multiple polylogarithms in characteristic 𝑝: 𝑣-adic vanishing versus ∞-adic Eulerianness, Int. Math. Res. Not. IMRN 2019 (2019), no. 3, 923–947. 10.1093/imrn/rnx151Search in Google Scholar

[10] C.-Y. Chang, M. A. Papanikolas and J. Yu, Frobenius difference equations and algebraic independence of zeta values in positive equal characteristic, Algebra Number Theory 5 (2011), 111–129. 10.2140/ant.2011.5.111Search in Google Scholar

[11] H.-J. Chen, On shuffle of double zeta values over F q ⁱ [ t ] , J. Number Theory 148 (2015), 153–163. 10.1016/j.jnt.2014.09.016Search in Google Scholar

[12] P. Deligne, Le groupe fondamental unipotent motivique de G m − ÎŒ N , pour N = 2 , 3 , 4 , 6 ou 8, Publ. Math. Inst. Hautes Études Sci. 112 (2010), 101–141. 10.1007/s10240-010-0027-6Search in Google Scholar

[13] P. Deligne and A. B. Goncharov, Groupes fondamentaux motiviques de Tate mixte, Ann. Sci. Éc. Norm. SupĂ©r. (4) 38 (2005), no. 1, 1–56. 10.1016/j.ansens.2004.11.001Search in Google Scholar

[14] C. Glanois, Motivic unipotent fundamental groupoid of G m ∖ ÎŒ N for N = 2 , 3 , 4 , 6 , 8 and Galois descents, J. Number Theory 160 (2016), 334–384. 10.1016/j.jnt.2015.08.003Search in Google Scholar

[15] A. B. Goncharov, The double logarithm and Manin’s complex for modular curves, Math. Res. Lett. 4 (1997), no. 5, 617–636. 10.4310/MRL.1997.v4.n5.a1Search in Google Scholar

[16] N. Green and T. Ngo Dac, On log-algebraic identities for Anderson 𝑡-modules and characteristic 𝑝 multiple zeta values, Int. Math. Res. Not. IMRN 2023 (2023), no. 16, 13687–13756. 10.1093/imrn/rnac141Search in Google Scholar

[17] R. Harada, Alternating multizeta values in positive characteristic, Math. Z. 298 (2021), no. 3–4, 1263–1291. Search in Google Scholar

[18] B. H. Im, H. Kim, K. N. Le, T. Ngo Dac and L. H. Pham, Zagier–Hoffman’s conjectures in positive characteristic, preprint (2022), https://arxiv.org/abs/2205.07165. Search in Google Scholar

[19] B. H. Im, H. Kim, K. N. Le, T. Ngo Dac and L. H. Pham, Hopf algebras and alternating multiple zeta values in positive characteristic, preprint (2023), https://arxiv.org/abs/2304.02337. Search in Google Scholar

[20] B. H. Im, H. Kim, K. N. Le, T. Ngo Dac and L. H. Pham, Hopf algebras and multiple zeta values in positive characteristic, preprint (2023), https://arxiv.org/abs/2301.05906. Search in Google Scholar

[21] B. H. Im, H. Kim, K. N. Le, T. Ngo Dac and L. H. Pham, Zagier–Hoffman’s conjectures in positive characteristic II, preprint (2024), https://arxiv.org/abs/2402.11539. Search in Google Scholar

[22] M. A. Papanikolas, Tannakian duality for Anderson–Drinfeld motives and algebraic independence of Carlitz logarithms, Invent. Math. 171 (2008), no. 1, 123–174. 10.1007/s00222-007-0073-ySearch in Google Scholar

[23] J. R. Schott, Matrix Analysis for Statistics, Wiley Ser. Probab. Stat., Wiley-Interscience, Hoboken, 2005. Search in Google Scholar

[24] D. S. Thakur, Function Field Arithmetic, World Scientific, River Edge, 2004. 10.1142/9789812562388Search in Google Scholar

[25] D. S. Thakur, Power sums with applications to multizeta and zeta zero distribution for F q ⁱ [ t ] , Finite Fields Appl. 15 (2009), no. 4, 534–552. 10.1016/j.ffa.2009.04.002Search in Google Scholar

[26] D. S. Thakur, Shuffle relations for function field multizeta values, Int. Math. Res. Not. IMRN 2010 (2010), no. 11, 1973–1980. Search in Google Scholar

[27] D. S. Thakur, Multizeta values for function fields: A survey, J. ThĂ©or. Nombres Bordeaux 29 (2017), no. 3, 997–1023. 10.5802/jtnb.1009Search in Google Scholar

[28] Y.-R. Yeo, Algebraic independence results for colored multizeta values in characteristic 𝑝, Int. J. Number Theory 19 (2023), no. 3, 677–708. 10.1142/S1793042123500343Search in Google Scholar

[29] D. Zagier, Values of zeta functions and their applications, First European Congress of Mathematics. Vol. II, Progr. Math. 120, BirkhĂ€user, Basel (1994), 497–512. 10.1007/978-3-0348-9112-7_23Search in Google Scholar

Received: 2023-06-15
Revised: 2024-01-23
Published Online: 2024-04-05
Published in Print: 2025-02-01

© 2024 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 11.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/forum-2023-0226/html
Scroll to top button