Preprint
Article

This version is not peer-reviewed.

Non-Archimedean Riesz-Frechet Representation

Submitted:

11 June 2026

Posted:

12 June 2026

You are already at the latest version

Abstract
We derive a Riesz-Frechet representation for bounded linear functionals defined on the padic Hilbert spaces introduced by Kalisch [Ann. of Math. (2), 1947]. We also notice the surprising difference between the Archimedean case and the non-Archimedean case (exact non-Archimedean version of Riesz-Frechet representation fails).
Keywords: 
;  ;  

1. Introduction

The century-old Riesz-Frechet representation (theorem) states that every bounded linear functional on a real/complex Hilbert space is completely determined by a single element in that Hilbert space.
Theorem 1.
[1,2,4,8,9,10,11,12,13,14,18,19,20,21,23,26,28,31,32,33,34,35,36,39,43,44,46,46] (Riesz-Frechet Representation) Let H be a complex Hilbert space. Let f : H C be a bounded linear functional. Then there is a unique element τ H satisfying the following conditions.
(i) 
f ( h ) = h , τ for all h H .
(ii) 
f = τ .
Over time, Riesz-Frechet representation has been extended to Lebesgue spaces [30,37,42] and spaces of continuous functionals on compact Hausdorff spaces (celebrated Riesz-Markov-Kakutani representation) [2,3,7,8,15,24,29,35,38,45]. It is known that Riesz-Frechet representation also holds for quaternionic Hilbert spaces [16]. Riesz-Frechet representation has also been derived for certain classes of Banach spaces by Giles [17] and James [22]. The traditional proof of the Riesz-Frechet representation uses the orthogonal decomposition theorem, which rests on the parallelogram law [19,32]. If the Hilbert space is separable, then one can also give a proof of the Riesz-Frechet representation for Hilbert spaces using orthonormal bases, which relies on the Gram-Schmidt orthonormalization process [20]. We are fundamentally motivated by the following question.
Question 2.
What is the non-Archimedean Riesz-Frechet representation?
We begin by recalling the definition of non-Archimedean Banach and p-adic Hilbert spaces.
Definition 1.
[40] Let K be a field. A map | · | : K [ 0 , ) is said to be a  non-Archimedean valuation  if the following conditions hold:
(i) 
If λ K satisfies | λ | = 0 , then λ = 0 .
(ii) 
| λ μ | = | λ | | μ | for all λ , μ K .
(iii) 
(Ultratriangle inequality) | λ + μ | max { λ | , | μ | } for all λ , μ K .
We say K isnon-Archimedean complete valued fieldif it is complete w.r.t. the ultrametric
d ( λ , μ ) | λ μ | , λ , μ K .
Definition 2.
[41] Let K be a non-Archimedean complete valued field. A vector space X over K is said to benon-Archimedean Banach spaceif there is a map (called as ultra-norm) · : X [ 0 , ) satisfying the following conditions:
(i) 
If x X is such that x = 0 , then x = 0 .
(ii) 
λ x = | λ | x for all λ K , for all x X .
(iii) 
(Ultranorm inequality) x + y max { x , y } for all x , y X .
(iv) 
X is complete w.r.t. the ultrametric
d ( x , y ) : = x y , x , y X .
Definition 3.
[25] Let K be a non-Archimedean valued field (with valuation | · | ) and X be a non-Archimedean Banach space (with norm · ) over K . We say that X is ap-adic Hilbert spaceif there is a map (called asp-adic inner product) · , · : X × X K satisfying the following conditions:
(i) 
If x X is such that x , y = 0 for all y X , then x = 0 .
(ii) 
x , y = y , x for all x , y X .
(iii) 
α x , y = α x , y for all α K , for all x , y X .
(iv) 
x + y , z = x , z + y , z for all x , y , z X .
(v)
| x , y | x y for all x , y X .
Definition 3 is more general than the definition of p-adic Hilbert spaces used by Kalisch, which imposes further conditions [25]. We note that the Definition 3 is different than the p-adic Hilbert spaces defined recently by Claussnitzer and Thom [6] where an exact Riesz-Frechet representation is known [5]. We derive two different versions of non-Archimedean Riesz-Frechet representation with different assumptions and different conclusions.

2. Non-Archimedean Riesz-Frechet Representation

In the paper, we use the following notion of Schauder basis for non-Archimedean Banach spaces and orthonormal basis for p-adic Hilbert spaces.
Definition 4.
Let X be a non-Archimedean Banach space over K . A collection { τ n } n in X is said to be a  Schauder basis  for X if for every x X , there is a unique sequence { α n ( x ) } n in K such that
x = n = 1 α n ( x ) τ n ,
series converges in the ultra-norm on X .
Definition 5.
Let X be a p-adic Hilbert space over K . A Schauder basis { τ n } n for X is said to be ap-adic orthonormal basisfor X if the following conditions hold:
(i) 
τ n 1 for all n N .
(ii) 
τ n , τ m = δ n , m for all n , m N .
Note that given a p-adic orthonormal basis { τ n } n for X , condition (ii) in Definition 5 says that
1 = | τ n , τ n | τ n 2 , n N .
By using condition (i) in Definition 5 we then get τ n = 1 for all n N . Thus all elements of p-adic orthonormal basis lie in the unit sphere. Following is the standard example of p-adic Hilbert space we keep in mind.
Example 1.
Let K be a non-Archimedean valued field. Define
c 0 ( N , K ) { a n } n : a n K , n N , lim n a n = 0 .
We define
{ a n } n max n N | a n | , { a n } n c 0 ( N , K )
and
{ a n } n , { b n } n n = 1 a n b n , { a n } n , { b n } n c 0 ( N , K ) .
Then c 0 ( N , K ) is a p-adic Hilbert space. The canonical Schauder basis { e n } n is a p-adic orthonormal basis for c 0 ( N , K ) .
Gram-Schmidt orthonormalization says that every separable Hilbert space has an orthonormal basis [27]. Eventhough we know from Example 1 that we have examples of p-adic orthonormal bases, we don’t know existence result of p-adic orthonormal basis in every p-adic Hilbert space.
Following is our first version of Riesz-Frechet representation.
Theorem 3.
(Non-Archimedean Riesz-Frechet Representation) Let X be a p-adic Hilbert space over a complete non-Archimedean valued field K . Let f : X K be a bounded linear functional. Assume that there is a p-adic orthonormal basis { ω n } n for X such that
lim n f ( ω n ) ω n = 0 .
Then there is a unique element τ X satisfying the following conditions:
(i) 
f ( x ) = x , τ for all x X .
(ii) 
f = τ .
Proof. 
Since { ω n } n is orthonormal,
k = 1 n f ( ω k ) ω k = max 1 k n | f ( ω k ) | , n N .
Since X is complete and lim n f ( ω n ) ω n = 0 , the series n = 1 f ( ω n ) ω n converges in X . Now define
τ n = 1 f ( ω n ) ω n .
We then have
f ( x ) = f n = 1 x , ω n ω n = n = 1 x , ω n f ( ω n ) = x , n = 1 f ( ω n ) ω n = x , τ , x X .
Now
f = sup | f ( x ) | x : x X , x 0 = sup | x , τ | x : x X , x 0 τ .
We now observe that
τ = n = 1 f ( ω n ) ω n = max n N | f ( ω n ) | f max n N ω n = f .
Uniqueness easily follows using the definition of p-adic inner product. □
Example 2.
Let p be a prime number. Define
f : c 0 ( N , Q p ) { a n } n f ( { a n } n ) n = 1 p n a n Q p .
Let { a n } n c 0 ( N , Q p ) . Then
| p n a n | p = 1 p n | a n | p 0 as n .
Hence
lim n p n a n = 0 , { a n } n c 0 ( N , Q p ) .
So the map f is well-defined. Define
τ p n n c 0 ( N , Q p ) .
Then
f ( x ) = x , τ , x c 0 ( N , Q p ) and f = τ = max n N | p n | p = max n N 1 p n = 1 p .
In the next example we show that every bounded linear functional on a p-adic Hilbert space cannot be represented by an element in the space.
Example 3.
Let p be a prime number. Define
f : c 0 ( N , Q p ) { a n } n f ( { a n } n ) n = 1 n a n Q p .
Let { a n } n c 0 ( N , Q p ) . Then
| n a n | p = | n | | a n | p | a n | p 0 as n .
Hence
lim n n a n = 0 , { a n } n c 0 ( N , Q p ) .
So the map f is well-defined. We claim that there is no τ c 0 ( N , Q p ) such that f ( x ) = x , τ for all x c 0 ( N , Q p ) . Let us suppose that this claim fails. Then thee exists a τ c 0 ( N , Q p ) such that f ( x ) = x , τ for all x c 0 ( N , Q p ) . Set τ { a n } n c 0 ( N , Q p ) . Let { e n } be the standard p-adic orthonormal basis for c 0 ( N , Q p ) . Now we see that
n = f ( e n ) = e n , τ = a n , n N .
But then τ = { n } n c 0 ( N , Q p ) because | n | p 0 as n . Hence Riesz-Frechet representation fails.
Let X be a p-adic Hilbert space. A closed subspace Y of X is said to be orthogonally complementable if there exists a unique closed subspace Z of X satisfying the following: for every x X , there exist unique y Y and z Z such that
(i)
x = y + z .
(ii)
y , z = 0 .
(iii)
x = max { y , z } .
In this case, we write x = y z , Z = Y and X = Y Y .
Following is our second version of Riesz-Frechet representation.
Theorem 4.
(Non-Archimedean Riesz-Frechet Representation) Let X be a p-adic Hilbert space over a complete non-Archimedean valued field K . Let f : X K be a non-zero bounded linear functional. Assume that Ker ( f ) is orthogonally complementable in X and there exists a (nonzero) vector ω Ker ( f ) such that ω , ω 0 . Then there is a unique element τ X satisfying the following conditions:
(i) 
f ( x ) = x , τ for all x X .
(ii) 
| τ , τ | τ f τ .
Moreover,
τ = | f ( ω ) | | ω , ω | ω f ω 2 | ω , ω | .
Proof. 
Define
τ f ( ω ) ω , ω ω .
Let x X . Then
f x f ( x ) f ( ω ) ω = f ( x ) f ( x ) = 0 .
Hence
x f ( x ) f ( ω ) ω Ker ( f ) .
We then have
0 = x f ( x ) f ( ω ) ω , τ = x , τ f ( x ) f ( ω ) ω , τ = x , τ f ( x ) f ( ω ) ω , τ = x , τ f ( x ) f ( ω ) ω , f ( ω ) ω , ω ω = x , τ f ( x ) .
Note that
| τ , τ | = | f ( τ ) | f τ .
We next extend the non-Archimedean Riesz-Frechet representation to non-Archimedean fields admitting conjugation (involution). We use the following definition of conjugation.
Definition 6.
Let K be a non-Archimedean complete valued field (with valuation | · | ). A map σ : K K is said to be aconjugationorinvolutionif the following conditions hold:
(i) 
σ ( α + β ) = σ ( α ) + σ ( β ) for all α , β K .
(ii) 
σ ( α β ) = σ ( α ) σ ( β ) for all α , β K .
(iii) 
σ ( σ ( α ) ) = α for all α K .
(iv) 
| σ ( α ) | = | α | for all α K .
In this case, we say that K is σ-conjugated.
Definition 7.
Let K be a σ-conjugated non-Archimedean complete valued field (with valuation | · | ) and X be a non-Archimedean Banach space (with norm · ) over K . We say that X is ap-adic σ -Hilbert spaceif there is a map (called asp-adic σ -inner product) · , · : X × X K satisfying the following:
1.
If x X is such that x , y = 0 for all y X , then x = 0 .
(i) 
x , y = σ ( y , x ) for all x , y X .
(ii) 
α x , y = α x , y for all α K , for all x , y X .
(iii) 
x + y , z = x , z + y , z for all x , y , z X .
(iv) 
| x , y | x y for all x , y X .
We now have p-adic versions of Theorem 3 and Theorem 4.
Theorem 5.
(Non-Archimedean Riesz-Frechet Representation) Let X be a p-adic σ-Hilbert space over a σ-conjugated complete non-Archimedean valued field K . Let f : X K be a bounded linear functional. Assume that there is a p-adic orthonormal basis { ω n } n for X such that
lim n f ( ω n ) ω n = 0 .
Then there is a unique element τ X satisfying the following conditions:
(i) 
f ( x ) = x , τ for all x X .
(ii) 
f = τ .
Proof. 
Define
τ n = 1 σ ( f ( ω n ) ) ω n .
The other arguments are similar to those used in the proof of Theorem 3. □
Theorem 6.
(Non-Archimedean Riesz-Frechet Representation) Let X be a p-adic σ-Hilbert space over a σ-conjugated complete non-Archimedean valued field K . Let f : X K be a non-zero bounded linear functional. Assume that Ker ( f ) is orthogonally complementable in X and there exists a (nonzero) vector ω Ker ( f ) such that ω , ω 0 . Then there is a unique element τ X satisfying the following conditions:
(i) 
f ( x ) = x , τ for all x X .
(ii) 
| τ , τ | τ f τ .
Moreover,
τ = | f ( ω ) | | ω , ω | ω f ω 2 | ω , ω | .
Proof. 
Define
τ σ ( f ( ω ) ) ω , ω ω .
The other arguments are similar to those used in the proof of Theorem 4. □

3. Conclusions

(1)
In 1907, Riesz [34] and Frechet [13] independently showed that every continuous linear functional on a Hilbert space is completely determined by a single element in that Hilbert space [4].
(2)
In 1938, Markov [29] and in 1941, Kakutani [24] extended the Riesz-Frechet representation to function spaces.
(3)
In 1947, James [22] and 1967, Giles [17] extended the Riesz-Frechet representation to certain classes of Banach spaces.
(4)
In this article, we derived two non-Archimedean Riesz representation theorems under certain conditions for p-adic Hilbert spaces.

References

  1. Akhiezer, N. I.; Glazman, I. M. Theory of linear operators in Hilbert space; Dover Publications: New York, NY, 1993. [Google Scholar]
  2. Charalambos D. Aliprantis and Kim C. Border. Infinite dimensional analysis: A hitchhiker’s guideSpringer, 2006.
  3. George Bachman and Lawrence Narici. Functional analysis. Dover Publications, New York, 2000.
  4. Garrett Birkhoff and Erwin Kreyszig. The establishment of functional analysis. Hist. Math. 1984, 11, 258–321. [CrossRef]
  5. Alcides Buss, Luiz Felipe Garcia, and Devarshi Mukherjee. Operator algebras over the p-adic integers. Trans. Am. Math. Soc. 2025, 378(10), 7371–7427.
  6. Anton Claussnitzer and Andreas Thom. Aspects of p-adic operator algebras. Münster J. Math. 2020, 13(2), 425–444.
  7. Donald L. Cohn. Measure theory. Birkhäuser Texts, Basler Lehrbüch. New York, NY: Birkhäuser/Springer, 2013.
  8. Conway, John B. A course in functional analysis; Springer-Verlag, 1990. [Google Scholar]
  9. Debnath, Lokenath; Mikusinski, Piotr. Introduction to Hilbert spaces with applications; Academic Press, 1999. [Google Scholar]
  10. Jean Dieudonne. History of functional analysis volume 49 of North-Holland Math. Stud. Elsevier, Amsterdam, 1981.
  11. Douglas, Ronald G. Banach algebra techniques in operator theory; Springer, 1998. [Google Scholar]
  12. Dunford, Nelson; Schwartz, Jacob T. Linear operators, I, General theory; Interscience Publishers, 1958. [Google Scholar]
  13. Frechet, M. Sur les operations lineaires. Trans. Am. Math. Soc. 1907, 8, 433–446. [Google Scholar] [CrossRef]
  14. Fuhrmann, Paul A. Linear systems and operators in Hilbert space; McGraw-Hill International Book Company, 1981. [Google Scholar]
  15. Garling, D. J. H. A ’short’ proof of the Riesz representation theorem. Proc. Camb. Philos. Soc. 1973, 73, 459–460. [Google Scholar] [CrossRef]
  16. Ghiloni, Riccardo; Moretti, Valter; Perotti, Alessandro. Continuous slice functional calculus in quaternionic Hilbert spaces. Rev. Math. Phys. 2013, 25(4), 83. [Google Scholar] [CrossRef]
  17. Giles, J. R. Classes of semi-inner-product spaces. Trans. Am. Math. Soc. 1967, 129, 436–446. [Google Scholar] [CrossRef]
  18. Gray, J. D. The shaping of the Riesz representation theorem: A chapter in the history of analysis. Arch. Hist. Exact. Sci. 1984, 31, 127–187. [Google Scholar] [CrossRef]
  19. Halmos, P. R. Introduction to Hilbert space and the theory of spectral multiplicity; Chelsea Publishing Company: New York, 1957. [Google Scholar]
  20. Halmos, Paul R. A Hilbert space problem book; Springer, 1982. [Google Scholar]
  21. Hansen, Vagn Lundsgaard. Functional analysis. Entering Hilbert space; World Scientific: Hackensack, NJ, 2016. [Google Scholar]
  22. James, Robert C. Orthogonality and linear functionals in normed linear spaces. Trans. Am. Math. Soc. 1947, 61, 285–292. [Google Scholar] [CrossRef]
  23. Kadison, Richard V.; Ringrose, John R. Fundamentals of the theory of operator algebras; Academic Press: San Diego, 1994; Vol. 1. [Google Scholar]
  24. Kakutani, Shizuo. Concrete representation of abstract (M)-spaces (A characterization of the space of continuous functions). Ann. Math. (2) 1941, 42, 994–1024. [Google Scholar] [CrossRef]
  25. Kalisch, G. K. On p-adic Hilbert spaces. Ann. Math. (2) 1947, 48, 180–192. [Google Scholar] [CrossRef]
  26. Kreyszig, Erwin. Introductory functional analysis with applications; John Wiley & Sons: New York, 1978. [Google Scholar]
  27. Leon, Steven J.; Bjorck, Ake; Gander, Walter. Gram-Schmidt orthogonalization: 100 years and more. Numer. Linear Algebra Appl. 2013, 20(3), 492–532. [Google Scholar]
  28. Limaye, Balmohan V. Functional analysis; New Age International Limited: New Delhi, 1996. [Google Scholar]
  29. Markov, A. A. On mean values and exterior densities. Rec. Math. Moscou N. Ser. 1938, 4, 165–191. [Google Scholar]
  30. McShane, E. J. Linear functionals on certain Banach spaces. Proc. Am. Math. Soc. 1950, 1, 402–408. [Google Scholar] [CrossRef]
  31. Muscat, Joseph. Functional analysis. An introduction to metric spaces, Hilbert spaces, and Banach algebras; Springer, 2024. [Google Scholar]
  32. Pavone, Marco. On the Riesz representation theorem for bounded linear functionals. Proc. R. Ir. Acad. Sect. A 1994, 94(1), 133–135. [Google Scholar]
  33. Pietsch, Albrecht. History of Banach spaces and linear operators; Birkhäuser: Basel, 2007. [Google Scholar]
  34. Riesz, F. Sur une espece de gometrie analytique des systemes de fonctions sommables. C. R. Acad. Sci. Paris 1907, 144, 1409–1411. [Google Scholar]
  35. Riesz, Friedrich. Zur Theorie des Hilbertschen Raumes. Acta Litt. Sci. Szeged 1934, 7, 34–38. [Google Scholar]
  36. Riesz, Frigyes; Sz.-Nagy, Bela. Functional analysis; Blackie & Son, Ltd: London-Glasgow, 1956. [Google Scholar]
  37. Royden, Halsey; Fitzpatrick, Patrick M. Real analysis; Prentice Hall: New York, 2010. [Google Scholar]
  38. Rudin, Walter. Real and complex analysis.; McGraw-Hill, 1987. [Google Scholar]
  39. Rudin, Walter. Functional analysis; McGraw-Hill, 1991. [Google Scholar]
  40. Schikhof, W. H. Ultrametric calculus. An introduction to p-adic analysis; Cambridge University Press: Cambridge, 1984. [Google Scholar]
  41. Schneider, Peter. Nonarchimedean functional analysis; Springer Monographs in Mathematics. Springer-Verlag: Berlin, 2002. [Google Scholar]
  42. Shioji, Naoki. Simple proofs of the uniform convexity of Lp and the Riesz representation theorem for Lp. Am. Math. Mon. 2018, 125(8), 733–738. [Google Scholar] [CrossRef]
  43. Marshall Harvey Stone. Linear transformations in Hilbert space and their applications to analysis volume 15 of Am. Math. Soc. American Mathematical Society (AMS), Providence, RI, 1932.
  44. Weidmann, Joachim. Linear operators in Hilbert spaces; Springer, 1980. [Google Scholar]
  45. Yeh, J. Real analysis. Theory of measure and integration; World Scientific, 2014. [Google Scholar]
  46. Young, Nicholas. An introduction to Hilbert space; Cambridge University Press, 1988. [Google Scholar]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

Subscribe

© 2026 MDPI (Basel, Switzerland) unless otherwise stated

Accessibility

Disclaimer

Terms of Use

Privacy Policy

Privacy Settings