Preprint
Article

This version is not peer-reviewed.

Revisiting Probabilistic Metric Spaces

Submitted:

04 September 2025

Posted:

05 September 2025

You are already at the latest version

Abstract

The field of probabilistic metric spaces has an intrinsic interest based on a blend of ideas drawn from metric space theory and probability theory. The goal of the present paper is to introduce and study new ideas in this field. In general terms, we investigate the following concepts: linearly ordered families of distances and associated continuity properties, geometric properties of distances, finite range weak probabilistic metric spaces, generalized Menger spaces, and a categorical framework for weak probabilistic metric spaces. Hopefully, the results will contribute to the foundations of the subject.

Keywords: 
;  ;  ;  ;  ;  ;  ;  ;  ;  ;  ;  ;  ;  

1. Introduction

1.1. Preliminary Remarks

The original and primary source on probabilistic metric spaces is [1]. Current research in the field appears to fall into three basic areas: (i) foundational topics ([2,3,4,5]), (ii) fixed point theory ([6,7,8]), and (iii) models for applications ([9,10]) with the primary emphasis being on area (ii). The admittedly lofty aim of the present paper is to rejuvenate the foundations of the subject by reexamining the original concepts before most research was devoted to Menger spaces and triangle functions.

1.2. Organization

The main body of the paper consists of three sections. Section two presents background material on weak probabilistic metric spaces (WPMSs) and introduces the notions of finite range WPMSs and linearly ordered families of distances. Section three presents results on constructing Menger spaces and approximating them by finite range Menger spaces, as well as characterizing the generalized Menger spaces. Section four introduces a categorical framework for WPMSs and establishes several key properties of the category WP. It also shows that the Menger spaces are a reflective subcategory of WP. Section five discusses the potential contributions of the work. There is also an appendix devoted to an overview of distance-spaces that includes several results used in other sections of the paper.

1.3. Conventions

We use the symbol § to refer to sections in the paper and to denote the end of proofs. The symbol N (resp. Z or R) denotes the natural numbers (resp. integers or real numbers), N+ = N\{0}, and R+ = [0, +∞). The symbol I denotes [0, 1]. For each xR, ⌊x⌋ (resp. ⌈x⌉) denotes the largest integer ≤ x (resp. smallest integer ≥ x). For a set S, idS : SS denotes the identity mapping, and for each AS, χA : S → {0, 1} (resp. inclA : AS) denotes the characteristic function of A (resp. the inclusion mapping). The symbol P(S) denotes the power-set of S and PF+(S) denotes the family of non-empty finite subsets of S. Given sets S and T, Sd T denotes their disjoint union, F(S, T) is the family of mappings ST, and for each tT, constt : ST is the constant mapping at t. A mapping φ : R+R+ is a modulus of continuity if φ is non-decreasing and φ(0+) = φ(0) = 0.

2. Weak Probabilistic Metric Spaces

2.1. Background

Here we present the notion of a weak probabilistic metric space, the associated family of distances, and several examples. The reader familiar with these ideas can skip to §2.2.
Define the set of distribution functions
Δ = {F : RI | F is left-continuous, non-decreasing, inf(F) = 0, and sup(F) = 1}
and let denote the point-wise order on Δ. The symbol H denotes the member of Δ defined by H(x) = 0 if x ≤ 0 and H(x) = 1 if x > 0. Let Δ+ = {F ∈ Δ | F(0) = 0}. If F ∈ Δ+, then F(x) = 0 for each x < 0, so we can assume that R+ is the domain of each member of Δ+.
A weak probabilistic metric space (WPMS) consists of a set S and F = {Fpq | p, qS} ⊆ Δ+ that satisfies the following properties:
(w1)  For each pS, Fpp = H.
(w2)  For each p, qS, Fpq = Fqp.
(w3)  For each distinct p, q, rS and x, yR+, Fpq(x) = Fqr(y) = 1 ⇒ Fpr(x + y) = 1.
The standard interpretation is that for p, qS and xR+, Fpq(x) is the probability that the distance from p to q is less than x. For example, Fpq(0) = 0 implies a zero probability that the distance between points p and q is less than 0. Property (w3) postulates a probabilistic triangle inequality: if the probability that the distance from p to q is less than x and the probability that the distance from q to r is less than y, then the probability that the distance from p to r is less than x + y. We note that (w3) also holds if p, q, and r are not distinct; for instance, if p = q, then Fpr(y) = 1, so since Fpr is non-decreasing, Fpr(x + y) = 1.
The property (w1) is a weaker version of the following property:
(w1′)  For each p, qS, Fpq = Hp = q.
Otherwise, our definition of a WPMS agrees with the one used in ([1], 1.4.2).
➤  In the literature, many authors use the phrase “probabilistic metric space” to mean a Menger space under a t-norm (see §3.1).
In the sequel, a pair (S, F) denotes a WPMS. Frequently, we also use the alternate representation (S, F), where F : S × S → Δ+ is defined by F(p, q) = Fpq.
We distinguish the following types of WPMSs. If (S, F) is a WPMS and (w1′) holds, then we say that (S, F) is a WPMS+. We say that a WPMS (S, F) is trivial if for each p, qS, Fpq = H. More generally, we say that a WPMS (S, F) is special if for each p, qS, Fpq-1(1) ≠ ∅.
Example 2.1.1.
The WPMS (S, G) determined by a pseudometric space (S, γ) ([1], 1.2.5).
For p, qS, define Gpq : R+I by Gpq(x) = H(x - γ(p, q)). Using the standard interpretation, the probability is zero that the distance from p to q is less than γ(p, q).
Since the motivation for a WPMS is to model a probabilistic situation, it is natural to use a distribution function to define the members of F. The next example illustrates this idea.
Example 2.1.2.
The 1-simple WPMS (S, A, γ, Φ) based on a pseudometric space (S, γ) and a fixed distribution function Φ ∈ Δ+\{H} ([1], 8.6.1). For p, qS, define Apq : R+I by
H(x) γ(p, q) = 0
Apq(x) = ⧼ 
⎩ Φ(x/γ(p, q)) γ(p, q) > 0.
If γ(p, q) = 0, then the probability is one that the distance from p to q is less than any positive number. If γ(p, q) > 0, then the probability that the distance from p to q is less than x is determined by the choice of Φ. In the same manner, we can define an α-simple WPMS (S, A, γ, Φ, α) for each 0 < α < 1 by using Φ(x/γ(p, q)α) in place of Φ(x/γ(p, q)).
Example 2.1.3.
For a distinct pair p, qS = R+, define Fpq : R+I by
⎧ 0 0 ≤ x < 1
Fpq(x) = ⧼  (x - 1)/apq 1 ≤ x < 1 + apq
⎩ 1 x ≥ 1 + apq
where apq = | p - q |. Also, for each pS, let Fpp = H. Then (R+, F) is a special WPMS.
Example 2.1.4.
For each p, qR satisfying | p - q | < 1, define fpq : R+I by
⎧ 0 x = 0
fpq(x) = ⧼  1 - | p - q | 0 < x ≤ 1
⎩ 1 x > 1.
Also, for each p, qR, define Fpq : R+I by
H(x - 1) | p - q | ≥ 1
Fpq(x) = ⧼ 
fpq(x) | p - q | < 1.
Then (R, F) is a WPMS+.
There is a natural way to associate distances with a WPMS. For each F ∈ Δ+, define the non-decreasing mapping F^: (0, 1] → [0, +∞] by F^(t) = sup(X(t)), where X(t) = {xR+ | F(x) < t} ([1], section 4.4). Since F(0) = 0, X(t) ≠ ∅ for each t > 0 and if t < 1, then X(t) is also bounded since sup(F) = 1, so F^(t) is well-defined. If F-1(1) ≠ ∅, then F^(1) is also well-defined.
Given a WPMS (S, F) and 0 < t < 1, define ωt : S × SR+ by
ωt(p, q) = Fpq^(t) = sup{xR+ | Fpq(x) < t}
and define the extended real-valued mapping ω1 : S × S → [0, +∞] by ω1(p, q) = Fpq^(1).
The ωt’s correspond to the dc’s introduced in ([1], 8.2.3). For each 0 < t < 1, ωt is a distance, i.e., for each p, qS, ωt(p, p) = 0 and ωt(p, q) = ωt(q, p) (§A.1). In general, ωt is not a semi-metric, that is, ωt(p, q) = 0 can occur for distinct p and q.
Given a WPMS (S, F), let
ΩF = {ωt | 0 < t < 1}
denote the set of well-defined distances. Notice that ΩF is a linearly ordered family of distances: if 0 < s < t < 1, then ωs ≤ ωt.
For an α-simple space (S, A, γ, Φ, α) (Example 2.1.2), ωt = Φ^(tα for each 0 < t ≤ 1 and in Example 2.1.3, ωt(p, q) = 1 + tapq for distinct p, qS and 0 < t < 1. Notice that in both cases, each ωt is a pseudometric. In the following example, the distances are not pseudometrics.
Example 2.1.5.
For each p, qR+, define Fpq : R+I by
Fpq(x) = [H(x - | p - q |) + H(x - | p2 - q2 |)]/2.
Then (R+, F) is a special WPMS+. As an illustration, if 0 < a = | p - q | < b = | p2 - q2 |, then
⎧ 0 0 ≤ xa
Fpq(x) = ⧼  1/2 a < xb
⎩ 1 x > b.
The distances fall into two groups:
e1(p, q) 0 < t ≤ 1/2
ωt(p, q) = ⧼ 
e2(p, q) 1/2 < t ≤ 1,
where e1 and e2 are defined by e1(p, q) = min{a, b} and e2(p, q) = max{a, b}, respectively. The mapping e2 is a metric, but e1 is not a pseudometric since the triangle inequality doesn’t hold. In fact, we can make a stronger statement. For nN+, let xn = n + 1/n, yn = n, and zn = -n + 1/n2. Then e1(xn, yn) + e1(yn, zn) = 3/n - 1/n4 → 0 as n → +∞, but e1(xn, zn) = 2 + 1/n2 + 2/n - 1/n4 > 2 for each n. Therefore, (R+, e1) doesn’t satisfy the condition (wtc) discussed in §3.3.

2.2. Preliminary Results

In this subsection, we present some basic results that provide further background for the reader as well as a basis for subsequent sections. In proofs, we often assume that p, q, rS and x, yR+ without explicitly mentioning the sets.
The first result summarizes the relationships between the values of the distribution functions and distances. We use it repeatedly in the text. The proof is elementary, but we include part of it to familiarize the reader with the notation.
Lemma 2.2.1.
The following statements hold for each WPMS (S, F) and 0 < t ≤ 1.
(a)
ωt(p, q) < xFpq(x) ≥ t.
(b)
Fpq(x) > t ⇒ ωt(p, q) < x.
(c)
Fpq(x) ≥ t ⇒ ωt(p, q) ≤ x and ωu(p, q) < x for each 0 < u < t.
Proof 
(b): Suppose y = ωt(p, q) ≥ x. By the left-continuity of Fpq, for each ε > 0, there is yε < y such that Fpq(y) - Fpq(yε) < ε. Hence, Fpq(yε) ≤ t, so Fpq(y) < t + ε. Since the inequality holds for each ε > 0, Fpq(y) ≤ t. Therefore, Fpq(x) ≤ t since Fpq is non-decreasing.
To check when a member of ΩF is a pseudometric, it is useful to introduce the following predicate. Given a WPMS (S, F) and 0 < t ≤ 1, define
(&t): ∀p, q, rS, ∀x, yR+Fpq(x) ≥ t and Fqr(y) ≥ tFpr(x + y) ≥ t.
It follows from property (w3) that (&1) always holds. The next result is implicit in the literature.
Prop 2.2.1.
Suppose (S, F) is a WPMS and 0 < t < 1.
(a)
If (&t) holds, then ωt is a pseudometric.
(b)
If ωs is a pseudometric for each 0 < s < t, then (&t) holds.
Proof 
(a): Let ε > 0 and choose x, y such that ωt(p, q) < x < ωt(p, q) + ε/2 and ωt(q, r) < y < ωt(q, r) + ε/2.
By Lemma 2.2.1(a), Fpq(x) ≥ t and Fqr(y) ≥ t, so by (&t), Fpr(x + y) ≥ t. Hence, by Lemma 2.2.1(c), ωt(p, r) ≤ x + y. Therefore, ωt(p, r) < ωt(p, q) + ωt(q, r) + ε for each ε > 0, so ωt satisfies the triangle inequality.
(b): If (&t) doesn’t hold, then there exists p, q, r, x, and y such that Fpq(x) ≥ Fqr(y) ≥ t > Fpr(x + y). Choose Fpr(x + y) < s < t. By Lemma 2.2.1(b), ωs(p, q) < x and ωs(q, r) < y. Since Fpr(x + y) < s, by Lemma 2.2.1(a), x + y ≤ ωs(p, r). Therefore, ωs(p, q) + ωs(q, r) < x + y ≤ ωs(p, r) shows that ωs doesn’t satisfy the triangle inequality.
The reader can see from the previous proof that Lemma 2.2.1 might need to be used repeatedly. Therefore, in future proofs, we only use it implicitly.
The following elementary result complements the preceding result.
Lemma 2.2.2.
The following statements are equivalent for a WPMS (S, F).
(a)
(S, F) is special.
(b)
ω1 is a real-valued mapping.
(c)
ω1 is a pseudometric.
Proof 
The implication (a) ⇒ (b) follows from the definition.
(b) ⇒ (c): Let x = ω1(p, q) and y = ω1(q, r). By definition, Fpq(x + ε/2) = 1 and Fqr(y + ε/2) = 1 for each ε > 0, so by (w3), Fpr(x + y + ε) = 1. Hence, ω1(p, r) ≤ x + y + ε for each ε > 0, so ω1 satisfies the triangle inequality.
(c) ⇒ (a): Since ω1(p, q) is finite, Fpq1(p, q) + 1) = 1. Hence, (S, F) is special.
The next two results present conditions that are equivalent to ΩF being a family of pseudometrics and metrics, respectively.
Prop 2.2.2.
The following statements are equivalent for a WPMS (S, F).
(a)
For each 0 < t < 1, ωt is a pseudometric.
(b)
For each 0 < t < 1, (&t) holds.
(c)
For each p, q, rS and x, yR+, Fpr(x + y) ≥ min{Fpq(x), Fqr(y)}.
Proof 
The implication (c) ⇒ (a) follows from Prop 2.2.1(a) since (c) implies that each (&t) holds. The implication (a) ⇒ (b) follows from Prop 2.2.1(b).
(b) ⇒ (c): Let s = Fpq(x) and t = Fqr(y). Clearly, if s = 0 or t = 0, then the inequality in (c) holds. Otherwise, if st, then by (&s), Fpr(x + y) ≥ s = min{s, t}. If s > t, then Fpr(x + y) ≥ t = min{s, t} follows from (&t).
The following result is not used in the rest of the paper, but it’s a natural companion to the previous proposition.
Prop 2.2.3.
The following statements are equivalent for a WPMS (S, F).
(a)
For each 0 < t < 1, ωt is a metric.
(b)
(i) For each pq, Fpq is right-continuous at 0.
(ii) For each p, q, rS and x, yR+, Fpr(x + y) ≥ min{Fpq(x), Fqr(y)}.
Proof 
(a) ⇒ (b): Suppose pq and let nN\{0, 1}. Since ω1/n(p, q) > 0, there is 0 < εn < 1/n such that εn < ω1/n(p, q). Hence, Fpqn) ≤ 1/n. Therefore, Fpq(0+) = limh→0+Fpq(h) = 0, which establishes (i). Condition (ii) follows from Prop 2.2.2.
(b) ⇒ (a): By Prop 2.2.2 and (ii), each ωt is a pseudometric. Suppose pq and ωt(p, q) = 0 for some 0 < t < 1. Then each Fpq(1/n) ≥ t, so Fpq is not right-continuous at 0, which contradicts (i).
We note that right continuity can be characterized by using the interplay between distances and distributions. Given a WPMS (S, F), for each xR+ and p, qS, define the predicate
(#x,p,q) ≡ 0 < t ≤ 1 and ωt(p, q) = xFpq(x) = t.
Then (S, F) satisfies (#x,p,q) if and only if Fpq is right-continuous at x. We leave this statement for the reader’s consideration.
The condition stated in ([5], (IV)) is equivalent to (&t) and the equivalence of parts (a) and (b) in Prop 2.2.2 and Prop 2.2.3 is proved in ([5], Lemma 1 and Theorem 1). Also, Prop 2.2.2(a) is proved in ([1], 8.2.3) and Prop 2.2.3 is essentially established in ([11], Theorems 9 and 10).

2.3. Reconstruction and Representation

The first result shows that for any WPMS (S, F), we can always reconstruct the family F = {Fpq} from ΩF.
PROP 2.3.1.
For each WPMS (S, F), p, qS, and xR+, Fpq(x) = sup{0 < t < 1 | ωt(p, q) < x}, where ΩF = {ωt}.
Proof 
For the proof, we suppress the pairs (p, q) and subscripts pq.
Let xR+ and T = {0 < t < 1 | ωt < x}. Define v = 0 if T = ∅ and v = sup(T) if T ≠ ∅. If T = ∅, then ωtx for each 0 < t < 1, so F(x) ≤ t for each 0 < t < 1. Hence, F(x) = v. If T ≠ ∅, then for each tT, F(x) ≥ t. Hence, vF(x). If v < F(x), choose v < r < F(x). Then ωr < x, so rT. Hence, vr, which is a contradiction. Therefore, v = F(x).
We remark that the following definition of ωt is used in ([5], (I)),: ωt(p, q) = inf{x | Fpq(x) > t}. Then ([5], (III)) asserts that Fpq(x) = sup{t | ωt(p, q) < x}.
Next, we present the first new idea in the paper. We say that a WPMS (S, F) has finite range if Rng(F) = ∪{Rng(Fpq) | p, qS} is a finite set. For this type of WPMS, we can establish the following representation theorem. Also, we’ll revisit these types of WPMSs in §3.2.
Theorem 2.3.1.
Assume that the WPMS (S, F) has the finite range Rng(F) = {ti | 0 ≤ in}, where t0 = 0 < t1 < … < tn = 1 and ΩF = {ωt}. Then the following statements hold:
(a)
ΩF = {ωt(i) | 1 ≤ in} and | ΩF | = n.
(b)
For each p, qS and xR+, Fpq(x) = Σ{(ti+1 - ti)H(x - ωt(i+1)(p, q)) | 0 ≤ i < n}.
Proof 
In various places, we write t(i) in place of ti.
(a): Given 0 < t ≤ 1, choose 1 ≤ in such that ti-1 < tti. Then for p, qS,
ωt(p, q) = sup{x | Fpq(x) < t} = sup{x | Fpq(x) < ti} = ωt(i)(p, q).
This establishes ωt = ωt(i). Let 1 ≤ i < n and choose p, q, and y such that Fpq(y) = ti. Then z = ωt(i+1)(p, q) ≥ y. Since Fpq is left-continuous, there exists 0 < x < y such that Fpq(x) = ti. If ωt(i)(p, q) ≥ z, then ωt(i)(p, q) > x, so Fpq(x) < ti, which gives a contradiction. Therefore, we have ωt(i)(p, q) < z = ωt(i+1)(p, q). It follows that | ΩF | = n.
(b): For each p, qS, define Gpq by Gpq(x) = Σ{(ti+1 - ti)H(x - ωt(i+1)(p, q)) | 0 ≤ i < n}.
(1)
(S, G) is a WPMS.
[Each Gpq ∈ Δ+ since H is left-continuous and Gpq(x) = 1 for x > ω1(p, q). It is easy to show that (w1) and (w2) hold. Suppose Gpq(x) = Gqr(y) = 1. Since (S, F) is special, by Lemma 2.2.2, ω1 is a pseudometric. Then by definition, x > ω1(p, q) and y > ω1(q, r), so x + y > ω1(p, r). Therefore, Gpr(x + y) = 1, so (w3) holds.]
Let ΩG = {νt}.
(2)
νt = ωt for each 0 < t ≤ 1.
[Let p, qS and choose 1 ≤ in such that ti-1 < tti. Since {ωt} is linearly ordered, by definition, Rng(G) ⊆ {ti | 0 ≤ in}. Hence, νt(p, q) = sup{x | Gpq(x) < t} = sup{x | Gpq(x) < ti} = νt(i)(p, q). By definition, if Gpq(x) ≤ ti-1, then x ≤ ωt(i)(p, q), so νt(i)(p, q) = sup{x | Gpq(x) < ti} ≤ ωt(i)(p, q). If Fpq(x) < ti, then x ≤ ωt(i)(p, q), so Gpq(x) < ti. Therefore,
ωt(i)(p, q) = sup{x | Fpq(x) < ti} ≤ sup{x | Gpq(x) < ti} = νt(i)(p, q).
This establishes ωt(i) = νt(i).]
Since (S, F) is a WPMS and by (1), (S, G) is a WPMS, it follows from (2) and Prop 2.3.1 that F = G.
The WPMS (S, F) in Example 2.1.5 illustrates Theorem 2.3.1 since
Fpq(x) = [H(x - | p - q |) + H(x - | p2 - q2 |)]/2 = [H(x - e1(a, b)) + H(x - e2(a, b))]/2.
In this case, | ΩF | = {e1, e2} and Rng(e1) = Rng(e2) = R+. Therefore, even for a WPMS with a finite range, the distances may not have finite ranges. In fact, Example 3.1.1 shows that the distances can have the values {0, 1} without the WPMS itself having a finite range.
The following result is an easy consequence of the previous theorem that fits nicely with one’s intuition. We can view it as a "0-1 law": if the probability is 0 or 1 that the distance from p to q is less than some value, then the distance is described by a unique pseudometric.
Corollary 2.3.1.
The following statements are equivalent for a WPMS (S, F).
(a)
Rng(F) = {0, 1}.
(b)
(S, F) is determined by a pseudometric.
(c)
ΩF = {ω1} and ω1 is a pseudometric.
(d)
| ΩF | = 1.
Proof 
(a) ⇒ (b): By Theorem 2.3.1, Fpq(x) = H(x - ω1(p, q)) for each p, qS and xR+. Since (S, F) is special, by Lemma 2.2.2, ω1 is a pseudometric.
(b) ⇒ (c): Suppose Fpq(x) = H(x - ω(p, q)) for a pseudometric ω and each p, qS and xR+. Then for each 0 < t ≤ 1, ωt(p, q) = sup{x | Fpq(x) < t} = sup{x | x ≤ ω(p, q)} = ω(p, q).
Clearly, (c) ⇒ (d).
¬(a) ⇒ ¬(d): If 0 < Fpq(x) < 1 for some p, qS and xR+, choose 0 < s < Fpq(x) < t < 1. Then ωs(p, q) < x ≤ ωt(p, q) shows that ωs ≠ ωt. Hence, | ΩF | > 1.

2.4. Linearly Ordered Families

Here, we present the second new idea in the paper – the application of linearly ordered families of distances. For clarity, we are referring to a family Ω = {ρt | 0 < t < 1} of distances defined on a set S that satisfies ρs ≤ ρt for each 0 < s < t < 1. We say that Ω satisfies the left-continuity property (lcp) (resp. right-continuity property (rcp)) if ρt = sup{ρs | 0 < s < t} (resp. ρt = inf{ρs | t < s < 1} for each 0 < t < 1.
Prop 2.4.1.
The following statements hold for a WPMS (S, F).
(a)
The family ΩF = {ωt | 0 < t < 1} satisfies (lcp) and if (S, F) is a special WPMS, then ω1 = supt | 0 < t < 1}.
(b)
The family ΩF satisfies (rcp) if and only if (#): for each distinct pair p, qS, Fpq is strictly increasing on Fpq-1(0, 1).
Proof 
(a): Let 0 < t < 1. For p, qS and ε > 0, there is xR+ such that ωt(p, q) - ε < x and Fpq(x) < t. If Fpq(x) < s < t, then ωs(p, q) ≥ x > ωt(p, q) - ε. Therefore, ωt = sups | 0 < s < t}. If (S, F) is special, then by Lemma 2.2.2, ω1 is real-valued, so by a similar argument, we can establish that ω1 = supt | 0 < t < 1}.
(b): Suppose ΩF satisfies (rcp) and p, qS is a distinct pair. Suppose 0 < Fpq(x) = Fpq(y) = t < 1 for some 0 ≤ x < y. If t < s < 1, ωs(p, q) ≥ y and ωt(p, q) ≤ x, so ωt(p, q) < infs(p, q) | t < s < 1}, which is a contradiction. Therefore, (#) holds. Conversely, if (#) holds and (rcp) doesn’t, then a = ωt(p, q) < b = infs(p, q) | t < s < 1} for some p and q. Then Fpq(x) ≥ t for each a < x and for each x < b and t < s < 1, Fpq(x) < s, so Fpq(x) ≤ t. Therefore, Fpq(x) = t for each a < x < b, which is a contradiction. Hence, (rcp) holds.
The following example describes linearly ordered families of pseudometrics that don’t satisfy the (lcp) and (rcp) conditions, respectively.
Example 2.4.1.
(1)
¬(lcp)
For each nN and 0 ≤ i < 2n, let
⎧ [i2-n, (i + 1)2-n) 0 ≤ i ≤ 2n - 2
Bi,n = ⧼ 
⎩ [1 - 2-n, 1] i = 2n - 1.
Each πn = {Bi,n | 0 ≤ i < 2n} is a partition of I and πn+1 < πn. Define the pseudometric dn on I by dn(p, q) = 0 if p, qBi,n for some i and dn(p, q) = 1 otherwise. For each 0 < t < 1, let ρt = dk, where k = ⌊t/(1 - t)⌋ and let ak = k/(k + 1) for kN. Then ρt = dk if and only if akt < ak+1. Suppose 0 < s < t < 1. If s ∈ [am, am+1) and t ∈ [an, an+1), where m < n, then ρs = dm and ρt = dn. Since πn < πm, ρs ≤ ρt, so Ω = {ρt} is a linearly ordered family. However, ρ1/2 = d1 is not the supremum of the family {ρt | 0 < t < 1/2} = {d0}, so (lcp) doesn’t hold.
(2)
¬(rcp)
For each distinct pair p, qI, define Fpq : R+I by
⎧ 0 x = 0
Fpq(x) = ⧼  min{p, q} 0 < x ≤ 1
⎩ 1 x > 1
and let Fpp = H. Then (I, F) is a WPMS and for each 0 < t ≤ 1,
⎧ 1 pq and min{p, q} < t.
ωt(p, q) = ⧼ 
⎩ 0 otherwise.
If p = 1/2 and q = 3/4, then Fpq = 1/2 on Fpq-1(0, 1), so by Prop 2.4.1(b), (rcp) doesn’t hold. Alternately, ω1/4(1/4, 3/4) = 0, but ωs(1/4, 3/4) = 1 for each 1/4 < s < 1, so (rcp) doesn’t hold.
The characterization stated in Prop 2.4.1(b) suggests another unexplored property. We say that a WPMS (S, F) satisfies the collective increasing property (cip) if for each 0 ≤ x < y, there exists p, qS such that Fpq(x) < Fpq(y). We can characterize this property by using the following set: given a WPMS (S, F), the spectrum of F is spec(F) = ∪{spect) | 0 < t < 1}, where ΩF = {ωt} and spect) = {ωt(p, q) | p, qS}.
The next result shows that the property (cip) is equivalent to a density criterion.
Prop 2.4.2.
A WPMS (S, F) satisfies (cip) if and only if spec(F) is a dense subset of R+.
Proof 
Suppose (S, F) satisfies (cip). Let 0 ≤ x < y and let x < a < b < y. By definition, there is p, qS such that Fpq(a) < Fpq(b). Choose Fpq(a) < t < Fpq(b). Then 0 < t < 1 and x < a ≤ ωt(p, q) < b, so spec(F) is dense in R+. Conversely, suppose spec(F) is dense in R+ and let 0 ≤ x < y. By assumption, there is 0 < t < 1 and p, qS such that x < ωt(p, q) < y. Then Fpq(x) < tFpq(y), so (cip) holds.
Because of the similar definitions, we can ask if (rcp) implies (cip). However, this isn’t the case. In Example 2.1.3, for distinct p and q, Fpq is strictly increasing on Fpq-1(0, 1) = (1, 1 + apq), so by Prop 2.4.1(b), (rcp) holds. Since each ωt(p, q) = 1 + tapq, spec(F) ⊆ {0} ∪ [1, +∞), so by Prop 2.4.2, (cip) doesn’t hold.
The next result shows that we can easily modify a linearly ordered family of pseudometrics so that (lcp) holds. We leave the proof to the reader.
Lemma 2.4.1.
Let Ω = {ρt | 0 < t < 1} be a linearly ordered family of distances (resp. pseudometrics) on a set S and for each 0 < t < 1, define ρ′t = sups | 0 < s < t}. Then Ω′ = {ρt′} is a linearly ordered family of distances (resp. pseudometrics) on S that satisfies (lcp) and ρt′ ≤ ρt for each t.
Other authors have shown that a WPMS can be reconstructed from its family of distances, but the following question doesn’t seem to have been addressed. What conditions on a family of distances guarantee that it has the form ΩF for some WPMS (S, F)? Based on Prop 2.4.1(a), (lcp) is a necessary condition and as we noted earlier, ΩF is a linearly ordered family. The next result shows how to construct a special WPMS from a linearly ordered family of distances. A companion result for Menger spaces is presented in §3.2.
Theorem 2.4.1.
Let Ω = {ρt | 0 < t ≤ 1} be a linearly ordered family of distances on a set S that satisfies the following conditions:
(a)
ρ1 is a pseudometric and ρ1 = supt | 0 < t < 1}.
(b)
If p, qS and ρ1(p, q) > 0, then ρt(p, q) < ρ1(p, q) for each 0 < t < 1.
Then there is a special WPMS (S, F) such that ω1 = ρ1 and ωt ≤ ρt for each 0 < t < 1, where ΩF = {ωt}. If Ω also satisfies (lcp), then ωt = ρt for each 0 < t < 1.
Proof 
For p, qS and xR+, let Tpq(x) = {0 < t < 1 | ρt(p, q) < x} and define Fpq : R+I by
⎧ 0 Tpq(x) = ∅
Fpq(x) = ⧼ 
sup(Tpq(x)) Tpq(x) ≠ ∅.
(1)
Each Fpq belongs to Δ+ and Fpq-1(1) ≠ ∅.
[Since Tpq(0) = ∅, Fpq(0) = 0. If x > 0, then Tpp(x) = (0, 1), so Fpp(x) = 1. Therefore, Fpp = H. Suppose p, qS is a distinct pair and let g = Fpq. It is routine to show that g(0) = 0 and g is non-decreasing. Also, by definition, g1(p, q) + 1) = 1, so sup(g) = 1. Let x > 0 and let {xn} be a sequence in (0, x) such that xnx. Let t = g(x). If t = 0, then g(xn) = 0 since g is non-decreasing, so limn →+∞g(xn) = g(x). Let t > 0 and choose 0 < ε < t. Since t = sup{0 < s < 1 | ρs(p, q) < x}, there exists t - ε < s < 1 such that ρs(p, q) < x. Then ρt-ε(p, q) ≤ ρs(p, q) < xn for sufficiently large n, so g(xn) ≥ t - ε. Hence, limn →+∞g(xn) = g(x). Therefore, g is left-continuous.]
(2)
(S, F) is a special WPMS.
[It’s routine to show that (w1) and (w2) hold. Suppose Fpq(x) = Fqr(y) = 1 for distinct p, q, rS and x, yR+. Then x > 0 and y > 0 and since Ω is linearly ordered, Tpq(x) = Tqr(y) = (0, 1). By (a), ρ1(p, q) = supt(p, q)} ≤ x and, similarly, ρ1(q, r) ≤ y, so ρ1(p, r) ≤ x + y. If ρ1(p, r) = 0, then each ρt(p, r) = 0, so Tpr(x + y) = (0, 1) since x + y > 0. If ρ1(p, r) > 0, then it follows from (b) that ρt(p, r) < ρ1(p, r) ≤ x + y for each 0 < t < 1, so Tpr(x + y) = (0, 1). It follows that in either case, Fpr(x + y) = 1, so (w3) holds. Then by (1), (S, F) is a special WPMS.]
Assume that ΩF = {ωt}. For the rest of the proof, we suppress the pair (p, q) and subscript pq.
(3)
For each 0 < t < 1, ωt ≤ ρt and ω1 = ρ1.
[Given ε > 0, choose x > ωt - ε such that F(x) < t. Then tT(x), so ρtx. Therefore, ωt ≤ ρt. If x > ω1, then F(x) = 1, so T(x) = (0, 1). Hence, ρt < x for each 0 < t < 1, so by (a), ρ1x. This establishes ρ1 ≤ ω1. By (2) and Prop 2.4.1(a), ω1 = supt} ≤ supt} = ρ1, so we have ω1 = ρ1.]
(4)
If Ω satisfies (lcp), then ωt = ρt for each 0 < t < 1.
[If ωt < ρt for some 0 < t < 1, then since (S, F) satisfies (lcp), there is 0 < u < t such that ωt < ρu. Hence, tFu), so Fu) = sup{0 < s < 1 | ρs < ρu} > 0. Since Ω is linearly ordered, if ρs < ρu, then s < u. Hence, Fu) ≤ u, which is a contradiction. Therefore, ωt = ρt.]
In some cases, we can replace condition (b) in the previous result without altering the conclusion. For example, if ρ1 satisfies (a) and (b′): ρ1(p, r) < ρ1(p, q) + ρ1(q, r) for distinct p, q, rS, then in the proof of statement (2), ρ1(p, r) < ρ1(p, q) + ρ1(q, r) ≤ x + y, so we obtain ρ1(p, r) < x + y. Therefore, Fpr(x + y) = 1.
Theorem 2.4.1 is suggested by Prop 2.3.1 and by the work on metrically generated spaces and E-spaces ([1], §1.7 and §9.2, [12]). The latter topic involves using a probability space and a family of metrics as a base space without referring to a linear order.
The following example illustrates the construction used in Theorem 2.4.1.
Example 2.4.2.
For each 0 < t ≤ 1, define the distance ρt on (0, 1) by ρt(p, q) = | p - q |2/(t+1). Then Ω = {ρt | 0 < t ≤ 1} satisfies the conditions in Theorem 2.4.1 and (lcp) holds, so ωt = ρt for each 0 < t ≤ 1. Now (S, F) is a special WPMS with the following distributions: for 0 < p < q < 1,
⎧ 0 x = 0
Fpq(x) = ⧼  2ln(q - p)/ln(x) - 1 0 < x < q - p
⎩ 1 xq - p.

3. Menger Spaces

In this section, we discuss the families of Menger spaces and generalized Menger spaces.

3.1. Definitions and Examples

Various conditions can be imposed on a WPMS (S, F) that reflect the properties found in standard examples. One well-known condition has the following form: there is a mapping T : I × II such that the following condition holds:
(T)  For each p, q, rS and x, yR+, Fpr(x + y) ≥ T(Fpq(x), Fqr(y)).
The mapping T is called a norm and other requirements are usually imposed. For instance, T is a t-norm ([1], 5.6.1) if it is commutative, associative, and satisfies the following conditions: (i) for aI, T(a, 1) = a and (ii) for a, b, c, dI, ac and bdT(a, b) ≤ T(c, d). Standard examples of t-norms are Min(a, b) = min{a, b}, Π(a, b) = ab, and Tm(a, b) = max{a + b - 1, 0} ([1], 11.1.1).
We say that a WPMS (S, F) is a Menger space under T ([1], 8.1.4) if T is a t-norm satisfying (T). If (S, F) is a Menger space under Min, then we simply refer to (S, F) as a Menger space. This is a non-standard use of the term that’s appropriate for our purposes. We note that if T is a t-norm satisfying (T), then TMin, so each Menger space is a Menger space under T.
Based on Prop 2.2.2, a WPMS (S, F) is a Menger space if and only if ΩF consists of pseudometrics; the sufficiency is noted [1], 8.2.3). The WPMSs in Examples 2.1.1 and 2.1.2 are Menger spaces. In Example 2.1.5, some distances are not pseudometrics, so the WPMS is not a Menger space. Here is another example of a Menger space.
Example 3.1.1.
For distinct p, qI, define Fpq : R+I by
⎧ 0 x = 0
Fpq(x) = ⧼  min{p, q} 0 < x ≤ 1
⎩ 1 x > 1.
Then (S, F) is a Menger space and for each 0 < t ≤ 1,
⎧ 1 pq and min{p, q} < t.
ωt(p, q) = ⧼ 
⎩ 0 otherwise.

3.2. Main Results

The first result is the promised companion to Theorem 2.4.1.
Theorem 3.2.1.
Let Ω = {ρt | 0 < t < 1} be a linearly ordered family of pseudometrics on a set S.
(a)
There is a Menger space (S, G) such that ωt ≤ ρt for each 0 < t < 1, where ΩG = {ωt}.
(b)
(S, G) is a special WPMSsupt(p, q) | 0 < t < 1} < +∞ for each p, qS. In this case, ρ1 is a pseudometric and ω1 ≤ ρ1 = supt}.
(c)
The family Ω satisfies (lcp) ⇔ ωt = ρt for 0 < t < 1. In this case, if (S, G) is special, then ω1 = ρ1.
Proof 
(a): For each p, qS and xR+, let Tpq(x) = {0 < t < 1 | ρt(p, q) < x} and define Gpq : R+I by
⎧ 0 Tpq(x) = ∅
Gpq(x) = ⧼ 
sup(Tpq(x)) Tpq(x) ≠ ∅.
(1)
Each Gpq belongs to Δ+.
[Since Gpqt(p, q) + 1) ≥ t for each 0 < t < 1, sup(Gpq) = 1. Now the rest of the proof used to establish statement (1) in Theorem 2.4.1 shows that Gpq belongs to Δ+.]
(2)
(S, G) is a Menger space.
[It’s routine to show that (w1) and (w2) hold. Suppose 0 < s = Gpq(x) ≤ t = Gqr(y) for p, q, rS and x, yR+. Given ε > 0, choose s - ε < s′ < s such that ρs′(p, q) < x and t - ε < t′ < t such that ρt′(q, r) < y. If s′ ≤ t′, then ρs(p, r) ≤ ρs(p, q) + ρt(q, r) < x + y, so Gpr(x + y) ≥ s′ > s - ε. Similarly, if t′ ≤ s′, then Gpr(x + y) ≥ t′ > t - ε ≥ s - ε. Hence, Gpr(x + y) ≥ s = min{Gpq(x), Gqr(y)}, so by (1), (S, G) is a Menger space.]
(3)
For each 0 < t < 1, ωt ≤ ρt.
[The same argument used to prove (3) in Theorem 2.4.1 establishes the statement.]
It follows from (2) and Prop 2.2.2 that each member of ΩG = {ωt} is a pseudometric.
(b): Let p, qS. If (S, G) is special, then Gpq(x) = 1 for some xR+, so supt(p, q)} ≤ x. Conversely, if y = supt(p, q)} < +∞, then Gpq(y + 1) = 1, so (S, G) is special. In addition, if supt(p, q)} < +∞, then by Prop 2.4.1(a), ρ1 = supt} is a pseudometric on S.
If Gpq(x) < 1 for p, qS and xR+, then ρ1(p, q) ≥ ρt(p, q) ≥ x for each Gpq(x) < t < 1, so ω1 ≤ ρ1.
(c): If ωt = ρt for each 0 < t < 1, then by (2) and Prop 2.4.1(a), Ω satisfies (lcp). Conversely, suppose Ω satisfies (lcp). If ωt(p, q) < ρt(p, q) for some 0 < t < 1 and p, qS, choose u < t that satisfies ωt(p, q) < x = ρu(p, q). Then tGpq(x). Since Ω is linearly ordered, if ρs(p, q) < x, then s < u, so Gpq(x) ≤ u. This is a contradiction, so ωt = ρt for each 0 < t < 1.
If (S, G) is special, then by Prop 2.4.1(a) and part (b), ω1 = supt} = supt} = ρ1.
Based on a summary of [13], a result similar to Theorem 3.2.1(b) seems to be established, but I can’t give an exact statement since I haven’t seen the paper. The following examples illustrate the construction used in the previous result.
Example 3.2.1.
In Example 2.4.1(1), for distinct p, qI, the distribution function in the corresponding Menger space is
⎧ 0 x = 0
Gpq(x) = ⧼  (n + 1)/(n + 2) 0 < x ≤ 1
⎩ 1 x > 1
where n is the largest integer such that p and q belong to the same member of πn. By Theorem 3.2.1(a), ωt ≤ ρt for each 0 < t < 1, where ΩG = {ωt}. Since Ω = {ρt} doesn’t satisfy (lcp) and ΩG does satisfy (lcp), ωt < ρt for some t. For example, ω1/2(0, 1/2) = 0 and ρ1/2(0, 1/2) = 1.
Example 3.2.2.
For each 0 < t < 1, define the distance ρt on I by ρt(p, q) = | p - q |1-t. Then Ω = {ρt} is a linearly ordered family of pseudometrics satisfying (lcp). If 0 < a = | p - q | < 1, then
⎧ 0 0 ≤ xa
Gpq(x) = ⧼  1 - ln(x)/ln(a) a < x < 1
⎩ 1 x ≥ 1
and
⎧ 0 0 ≤ x ≤ 1
G01(x) = ⧼ 
⎩ 1 x > 1.
The WPMS (I, G) is the Menger space referred to in Theorem 3.2.1 that satisfies Ω = ΩG.
In Theorem 2.3.1, we presented a representation result for WPMSs with finite range. Here, we show that the Menger spaces satisfying this condition can be used to approximate any Menger space. We begin with the following preliminary result.
Lemma 3.2.1.
Let (S, F) be a WPMS and V = {v0, …, vn}, where v0 = 0 < v1 < … < vn = 1. For each p, qS, define FVpq : R+I by FVpq(x) = min{vV | Fpq(x) ≤ v}. Then the following statements hold:
(a)
For p, qS, FVpq ∈ Δ+ and FVpq-1(1) ≠ ∅.
(b)
(S, FV) satisfies (w1) and (w2).
(c)
For p, qS, FpqFVpq and || FVpq - Fpq ||max{vk+1 - vk | 0 ≤ k < n}.
(d)
For 0 < t ≤ 1, define ωVt : S × SR+ by
ωVt(p, q) = sup{xR+ | FVpq(x) < t}.
Then ωVt is a distance and ωV1 is a pseudometric.
Proof 
Let p, qS and let G = FVpq. For xR+, let A(p, q, x) = {vV | Fpq(x) ≤ v}.
(a): Since 1 ∈ V, G is well-defined. Since Fpq(0) = 0, G(0) = 0. Since sup(Fpq) = 1, there is x such that Fpq(x) > vn-1, so G(x) = 1. Since Fpq is non-decreasing, if xy, then A(p, q, y) ⊆ A(p, q, x), so G(x) ≤ G(y). Let x > 0 and suppose G(x) = vk. If k = 0, then G(x) = 0, so G(p) = 0 for 0 < p < x. Hence, G is left-continuous at x. If k > 0, then vk-1 < Fpq(x) ≤ vk. Since Fpq is left-continuous, there exists 0 < h < x such that vk-1 < Fpq(x - h). Then G(p) = vk for each x - h < p < x, so G is left-continuous at x. This shows that G ∈ Δ+.
(b): If p = q and x > 0, then by (w1) for (S, F), G(x) = min(A(p, p, x)) = min{vV | H(x) ≤ v} = 1. Hence, since G(0) = 0, G = H, so (w1) holds. Since (S, F) satisfies (w2), for p, qS and xR+, A(p, q, x) = A(q, p, x), so FVpq = FVqp. Hence, (w2) holds.
(c): If G(x) = vk, then Fpq(x) ≤ vk, so FpqG. If k = 0, then G(x) = 0 = Fpq(x). Otherwise, k > 0, so vk-1 < Fpq(x) and G(x) - Fpq(x) < vk - vk-1. Hence, || FVpq - Fpq ||max{vk+1 - vk}.
(d): It follows from part (b) that each ωVt is a distance. Also, by part (a), ωV1 is well-defined, so by Lemma 2.2.2, it is a pseudometric.
We say that a WPMS (S, F) can be approximated by a family W of WPMSs if for each ε > 0, there exists (S, G) ∈ W such || Fpq - Gpq || < ε for each p, qS. Now we can establish the following result.
Theorem 3.2.2.
Each Menger space can be approximated by the family of special Menger spaces with a finite range.
Proof 
Let (S, F) be a Menger space and suppose V = {v0, …, vn} with v0 = 0 < v1 < … < vn = 1.
(1)
(S, FV) is a special Menger space with a finite range.
[Let p, q, rS, x, yR+, G = FVpq, and H = FVqr. Suppose G(x) = vk and H(y) = vm with km. If k = 0, then min{G(x), H(y)} = 0. If k > 0, then G(x) > vk-1 and H(y) > vm-1, so Fpq(x + y) > vk-1. Hence, FVpq(x + y) ≥ vk = min{G(x), H(y)}. A similar argument works if m < k. Therefore, (S, FV) satisfies condition (T) for T = Min, so by Lemma 3.2.1(a)(b), it is a special Menger space.
Given ε > 0, let V = {k/n | 0 ≤ kn}, where nN+ satisfies 1/n < ε. Therefore, for each p, qS, || FVpq - Fpq || ≤ 1/n by Lemma 3.2.1(c).
To establish Theorem 3.2.2, it appears that (S, F) must be a Menger space. Otherwise, there exists p, q, rS and x, yR+ such that a = Fpr(x + y) < b = Fpq(x) ≤ Fqr(y). Let V = {0, (a + b)/2, 1}. Then FVpq(x) = FVqr(y) = 1, but FVpr(x + y) ≤ (a + b)/2, so (S, FV) doesn’t satisfy (w3).

3.3. Geometric Conditions

The Menger spaces are characterized among WPMSs by the fact that each associated distance is a pseudometric. This is the gold standard. However, before the definition of a metric space was codefied, other properties of distances were studied ([14,15,16,17]). In particular one largely forgotten property of a distance-space (S, d) is the following weak triangle condition:
(wtc) ≡ ∀ε > 0 ∃δ > 0 ∀x, y, zMd(x, z) < δ and d(z, y) < δ ⇒ d(x, y) < ε.
Clearly, each pseudometric satisfies (wtc), but in Example 2.1.5, some distances do not satisfy it. Also, in Example 4.2.1(1), (R, F) is a WPMS+ and for each 0 < t ≤ 1,
⎧ 0 | p - q | ≤ 1 - t
ωt(p, q) = ⧼ 
⎩ 1 | p - q | > 1 - t.
Given 0 < t < 1, let p = 0, q = 1 - t, and r = 2(1 - t). Then ωt(p, q) = ωt(q, r) = 0, but ωt(p, r) = 1. Therefore, no member of ΩF satisfies (wtc). On the other hand, in Example 2.4.2, the (wtc) property holds for each distance.
The importance of the (wtc) property is demonstrated by Theorem A.3.1. If a distance-space (S, d) satisfies (wtc), then d is uniformly equivalent to a pseudometric on S. More generally, other special properties of distances may be useful for the classification of WPMSs. For purposes of comparison, we’ve included a second (wpc) property in §A.3.

3.4. Generalized Menger Spaces

Now we present the third new idea in the paper. Much study has been devoted to the study of Menger spaces, but it seems that a natural generalizations has not received much attention. Consider a WPMS that satisfies the following weaker version of (T), where T = Min:
(M′) ∀ε > 0 ∃δ > 0 ∀p, q, rSFpr(ε) ≥ min{Fpq(δ), Fqr(δ)}.
In this case, we say that (S, F) is a generalized Menger space. If (S, F) is a Menger space, then (M′) holds since Fpr(ε) ≥ min{Fpq(ε/2), Fqr(ε/2)} for each ε > 0.
The next result gives several characterizations of generalized Menger spaces.
Theorem 3.4.1.
The following conditions are equivalent for a WPMS (S, F).
(a)
(M′) holds.
(b)
There is a right-continuous modulus of continuity φ such that for p, q, rS and 0 < t < 1,
ωt(p, r) < φ(maxt(p, q), ωt(q, r)}).
(c)
There is a right-continuous modulus of continuity φ such that Fpr ° φ ≥ min{Fpq, Fqr} for p, q, rS.
(d)
ΩF satisfies (equi-wtc): for each ε > 0, there exists δ > 0 such that for each 0 < t < 1 and p, q, rS, ωt(p, r) < δ and ωt(q, r) < δ ⇒ ωt(p, q) < ε.
Proof 
(a) ⇒ (b): In the definition of (M′), δ depends only on ε, so the proof of (a) ⇒ (b) in Theorem A.3.1 shows that there is a non-decreasing mapping τ: (0, +∞) → (0, +∞) that satisfies the following conditions:
(ec1)  For each p, q, rS and 0 < t < 1, if ωt(p, q) < x and ωt(q, r) < x, then ωt(p, r) < τ(x).
(ec2)  τ(0+) = 0.
Define the mapping φ: R+R+ by φ(x) = τ(x+). Since τ is non-decreasing, φ is non-decreasing and an elementary argument shows that φ is right-continuous. Hence, by (ec2), φ(0+) = φ(0) = 0, so φ is a modulus of continuity that satisfies (ec1). Let p, q, rS, 0 < t < 1, and assume that ωt(p, q) ≤ ωt(q, r). By (ec1), if ωt(q, r) < x, then ωt(p, r) < φ(x). Hence, by the right-continuity of φ, ωt(p, r) ≤ φ(ωt(q, r)).
(b) ⇒ (c): Let xR+ and assume that s = Fpq(x) ≤ t = Fqr(x). Then ωs(q, r) ≤ ωt(q, r) ≤ x and ωs(p, q) ≤ x, so ωs(p, r) < φ(maxs(p, q), ωs(q, r)}) ≤ φ(x). Hence, Fpr(φ(x)) ≥ s.
(c) ⇒ (d): Since φ(0+) = φ(0) = 0, given ε > 0, there is δ > 0 such that φ(δ) < ε/2. Suppose ωt(p, q) < δ and ωt(q, r) < δ for some 0 < t < 1. Then Fpq(δ) ≥ t and Fqr(δ) ≥ t, so by (c), we obtain Fpr(ε/2) ≥ Fpr(φ(δ)) ≥ t. Hence, ωt(p, r) ≤ ε/2 < ε. Therefore, ΩF satisfies (equi-wtc).
(d) ⇒ (a): Let ε > 0 and choose δ > 0 such that the conclusion in (d) holds. Let p, q, rS and suppose α = Fpq(δ/2) ≤ β = Fqr(δ/2). If α = 0, then Fpr(ε) ≥ min{α, β}. If α = 1, then β = 1. Let 0 < t < 1. Then Fpq(δ/2) = Fqr(δ/2) > t, so ωt(q, r) < δ and ωt(q, r) < δ. Hence, ωt(p, r) < ε, so Fpr(ε) ≥ t. Since t is arbitrary, Fpr(ε) = 1, so Fpr(ε) ≥ min{α, β}. If 0 < α < 1, then ωα(p, q) < δ and ωα(q, r) < δ, so ωα(p, r) < ε. Hence, Fpr(ε) ≥ α= min{α, β}. This establishes (M′).
In Example 2.1.5, ωt doesn’t satisfy (wtc) for 0 < t ≤ 1/2, so by the previous result, the WPMS is not a generalized Menger space. The following example shows that a generalized Menger space may not be a Menger space.
Example 3.4.1.
We reuse Example 2.4.2. For 0 < t ≤ 1, the distance ρt on (0, 1) is defined by ρt(p, q) = | p - q |2/(t+1) and (S, F) is a special WPMS.
We claim that Ω = {ρt} is an (equi-wtc) family. For 0 < ε < 1, let δ = ε/4 and let p, q, and r be distinct points in (0, 1). Let a = | p - q |, b = | q - r |, and c = | p - r |. For 0 < t < 1, let y = 2/(t + 1), and suppose ay < δ and by < δ. If c = a + b, then cy < (2δ1/y)y = 2yδ= 2y-2ε < ε. If c = a - b, then cy < ay < ε and, similarly, if c = b - a, then cy < by < ε. By Theorem 3.4.1, (S, F) satisfies (M′). Choose 0 < t < 1 and let p = 1/4, q = 1/2, r = 3/4, and y = 2/(t + 1). Then ρt(p, q) = ρt(q, r) = (1/4)y, so ρt(p, q) + ρt(q, r) = 2(1/4)y = 21-2y and ρt(p, r) = 2-y. Since y > 1, 2-y> 21-2y, so ρt is not a pseudometric. Hence, by Prop 2.2.2, (S, F) is not a Menger space.
The following result shows that each generalized Menger space has a closely associated Menger space.
Theorem 3.4.2.
If a WPMS (S, F) satisfies (M′), then there is a Menger space (S, G) and right-continuous moduli of continuity f and g such that the following statements hold for p, qS and 0 < t < 1, where ΩF = {ωt} and ΩG = {σt}:
(a)
σt(p, q) < δ < 1/8 ⇒ ωt(p, q) < f(δ).
(b)
ωt(p, q) < δ < 1 ⇒ σt(p, q) < g(δ).
In terms of distribution functions, for 0 < δ < 1/8, Gpq(δ) ≤ Fpq(f(δ)) and Fpq(δ) ≤ Gpq(g(δ)).
Proof 
The proof is essentially the same as the proof of (b) ⇒ (c) in Theorem A.3.1 except that we use the regular écart φ from Theorem 3.4.1. In statements (1) and (2), ω denotes one of the mappings ωt. Their proofs follow the proof of Theorem A.3.1. Recall from §A.1 that if d is a distance on S, then S(d, r) = Sd(r) = {Sd(x, r) | xS} for each r > 0, where Sd(x, r) = {yS | d(x, y) < r}.
(1)
For each xM and r > 0, St(x, Sω(r)) ⊆ Sω(x, φ(r)).
(2)
There is a decreasing sequence {εk} ⊆ (0, 1] converging to 0 with ε0 = ε1 = 1 such that φ(εk+1) < εk and Sωk+1) <* Sωk) for each k > 0.
For each 0 < t < 1, let A0,t = {S} and Ak,t = S(ωt, εk) for k > 0. By (2), each {Ak,t} is a normal sequence. Define μt : S × SR by μt(p, q) = inf{2-k | qSt(p, Ak,t)} and let ρt be the corresponding path-metric based on Prop A.2.2(i). If 0 < s < t < 1, then Ak,t < Ak,s for each k since ωs ≤ ωt. Therefore, μs ≤ μt, so ρs ≤ ρt.
Define f : R+R+ by
⎧ 0 x = 0
f(x) = ⧼  φ(εk) 0 < x < 1/4 and k = ⌈-lg(4x)⌉
⎩ φ(1) x ≥ 1/4
and define h : R+N+ by
max{k | x < εk} 0 < x < 1
h(x) = ⧼ 
⎩ 1 x ≥ 1.
Define g : R+R+ by g(0) = 0 and g(x) = x + 2-h(x) if x > 0. One can verify that f and g are right-continuous moduli of continuity.
(3)
(i)   ρt(p, q) < δ < 1/4 ⇒ ωt(p, q) < f(δ).
(ii)   ωt(p, q) < δ < 1 ⇒ ρt(p, q) < g(δ).
[(i): By Prop A.2.2(ii), ρt ≤ μt ≤ 2ρt, so we obtain μt(p, q) < 2δ < 1. Let k = ‸-lg(2δ)‹. Then 2-k-1 ≤ 2δ < 2-k, so μt(p, q) ≤ 2-k. Since qSt(p, Ak,t) = St(p, Sω(t)k)), it follows from (1) that ωt(p, q) < φ(εk) = f(δ).
(ii): Let k = h(δ). Since δ < εk, it follows that ωt(p, q) < εk, so we have qSt(p, Ak,t). Therefore, ρt(p, q) ≤ μt(p, q) ≤ 2-k < g(δ).]
For each 0 < t < 1, let μ′t = sups | 0 < s < t} and ρ′t = sups | 0 < s < t}. Then μ′t ≤ μt, ρ′t ≤ ρt, and by Lemma 2.4.1, {ρ′t} is a linearly ordered family of pseudometrics satisfying (lcp). Therefore, by Theorem 3.2.1(a)(c), there is a Menger space (S, G) with σt = ρ′t for each t, where ΩG = {σt}.
(4)
For each 0 < t < 1, μt ≤ 2μ′t and ρt ≤ 2ρ′t.
[Assume that μ′t(p, q) = 2-k and choose 0 < s′ < t such that μs′(p, q) > 2-k-1. Let s′ < s < t. Then μs(p, q) = 2-k, so qSt(p, Ak,s). Hence, by (1), ωs(p, q) ≤ φ(εk). By Prop 2.4.1(a), the family {ωt} satisfies (lcp), so by (2), ωt(p, q) ≤ φ(εk) < εk-1. Therefore, μt(p, q) ≤ 21-k = 2μ′t(p, q). This establishes the first statement. The second statement follows from the first one and the definition of a path-metric.]
(5)
(i) σt(p, q) < δ < 1/8 ⇒ ωt(p, q) < f(δ)
(ii) ωt(p, q) < δ < 1 ⇒ σt(p, q) < g(δ).
[(i): By (4), ρt(p, q) ≤ 2σt(p, q) < 2δ < 1/4, so by (3)(i), ωt(p, q) < f(δ).
(ii): By (3)(ii), σt(p, q) ≤ ρt(p, q) < g(δ).]
To verify the final statement, assume that Gpq(δ) > Fpq(f(δ)) and choose Gpq(δ) > t > Fpq(f(δ)). Then σt(p, q) < δ, so by (5)(i), ωt(p, q) < f(δ). Hence, Fpq(f(δ)) ≥ t, which is a contradiction. Therefore, Gpq(δ) ≤ Fpq(f(δ)). The other statement is established similarly.]

4. Operations on WPMSs

4.1. The Category WP

In this section, we introduce the fourth new idea in the paper – a categorical framework for WPMSs and accompanying operations. The idea has been touched on in [4] and [18], where a generalized notion of probabilistic metric spaces is introduced in the context of quantales.
The reader is referred to [19] for general categorical notions. Let | WP | denote the family of WPMSs. For (S, F) and (T, G) in | WP |, a morphism [f]: (S, F) → (T, G) is a mapping f : ST such that Fpq  Gf(p)f(q) for each p, qS, where denotes the point-wise order on Δ+. The symbol WP((S, F), (T, G)) denotes the family of morphisms from (S, F) to (T, G). The composition of morphisms [f]: (S, F) → (T, G) and [g]: (T, G) → (V, K) is the morphism [g ° f ]: (S, F) → (V, K). Also, [idS] is the identity morphism for each WPMS (S, F), so WP is a category.
The first result characterizes the morphisms in WP in terms of the families of distances.
Lemma 4.1.1.
The following statements are equivalent for WPMSs (S, F) and (T, G) and a mapping f : ST, where ΩF = {ωt} and ΩG = {ρt}.
(a)
[f] : (S, F) → (T, G) is a WP-morphism [f] : (S, F) → (T, G).
(b)
For each 0 < t < 1, f : (S, ωt) → (T, ρt) is a non-expansive mapping.
Proof 
¬(b) ⇒ ¬(a): Choose t, p, and q such that ωt(p, q) < ρt(f(p), f(q)). Let ωt(p, q) < x < ρt(f(p), f(q)). Therefore, Gf(p)f(q)(x) < tFpq(x), so ¬(a) holds.
¬(a) ⇒ ¬(b): Choose p, q, and x such that Gf(p)f(q)(x) < Fpq(x) and let Gf(p)f(q)(x) < t < Fpq(x). Then ωt(p, q) < x ≤ ρt(f(p), f(q)). Hence, ωt(p, q) < ρt(f(p), f(q)), so ¬(b) holds.
Corollary 4.1.1.
The following statements are equivalent for WPMSs (S, F) and (T, G) and a mapping h : ST, where ΩF = {ωt} and ΩG = {ρt}.
(a)
[h] : (S, F) → (T, G) is a WP-isomorphism.
(b)
h is a bijection and Fpq = Gh(p)h(q) for each p, qS.
(c)
For each 0 < t < 1, h : (S, ωt) → (T, ρt) is a surjective isometry.
Proof 
The equivalence of (a) and (b) follows from the definition of a WP-isomorphism. The equivalence of (a) and (c) follows from Lemma 4.1.1.
We note that the notion of isomorphic WPMS’s is the same as the idea of isometric WPMS’s found in ([1], 8.1.2).
The next result lists some elementary facts about the category WP. The proof is left to the reader.
Lemma 4.1.2.
The following statements hold for WPMSs (S, F) and (T, G).
(a)
The monomorphisms (resp. epimorphisms) in WP are exactly the morphisms that are injections (resp. surjections) on the base sets.
(b)
For each tT, constt : ST defines a WP-morphism [constt] : (S, F) → (T, G).
(c)
Let HS and FH = F|H×H. Then (H, FH) is a WPMS and inclH : HS defines a WP-morphism [inclH] : (H, FH) → (S, F).
In the literature, both countable products and arbitrary products of WPMSs have been studied. In our setting, we can characterize when products exist.
For a set W = {(Si, Fi) | iI} ⊆ | WP |, let S = ΠSi and define the mapping F : S × SF(R+, I) by F(p, q) = inf{Fi(pi, qi) | iI}. Then the following statements hold:
(1)
For each p, qS, Fpq(0) = 0, Rng(Fpq) ⊆ I, and Fpq is non-decreasing. (This follows since each Fi(pi, qi) ∈ ρ+.)
(2)
(S, F) satisfies (w1) – (w3). (Clearly, (w1) and (w2) hold. If Fpq(x) = Fqr(y) = 1 for some p, q, rS and x, yR+, then each Fi(pi, qi)(x) = Fi(qi, ri)(y) = 1, so Fi(pi, ri)(x + y) = 1. Hence, Fpq(x + y) = 1, so (w3) holds.)
(3)
For each 0 < t < 1, ωt = supi,t}, where ΩF = {ωt} and each ΩF(i) = {ωi,t}.
It may not be the case that sup(Fpq) = 1 for each p, qS (Example 4.1.1), so Fpq may not belong to ρ+. To remedy this problem, we introduce the following property.
We say that W satisfies the uniform limit property (ulp) if for each ε > 0 and each {pi, qi} ⊆ Si, iI, there is xR+ such that inf{Fi(pi, qi)(x) | iI} > 1 - ε. Clearly, (ulp) holds if W is a finite family. Also, if W satisfies (ulp), then each Fpq∈ ρ+, so by (1) and (2), (S, F) is a WPMS.
Prop 4.1.1.
The following statements hold for a set W = {(Si, Fi) | iI} ⊆ | WP |.
(a)
If W satisfies (ulp), then (S, F) is the product in WP of the members of W. Hence, each finite family of WPMSs has a product in WP.
(b)
If W satisfies (ulp) and W consists of Menger spaces, then (S, F) is a Menger space.
(c)
If W is an infinite set and its product in WP exists, then W satisfies (ulp).
Proof 
(a): Based on our earlier discussion, (S, F) is a WPMS, where S = ΠSi and F : S × S → Δ+ is defined by F(p, q) = inf{Fi(pi, qi) | iI}. For each iI, let πi : SSi be the standard projection mapping. Since FFi, [πi] : (S, F) → (Si, Fi) is a morphism. Let (T, G) be a WPMS and assume that each [fi] : (T, G) → (Si, Fi) is a morphism. Let f : TS be the mapping that satisfies πi ° f = fi for each i. Then G(a, b) ≤ Fi(fi(a), fi(b)) for each a, bT and i, so G(a, b) ≤ F(f(a), f(b)). Therefore, [f] : (T, G) → (S, F) is a morphism. If the morphism [g] : (T, G) → (S, F) satisfies πi ° g = fi for each i, then clearly, f = g. Hence, (S, F) is the product.
(b): We refer to (2) above. Suppose F(p, r)(x + y) < b = F(p, q)(x) ≤ c = F(q, r)(y) for p, q, rS and x, yR+. Choose iI with Fi(pi, ri)(x + y) < b. Since bFi(pi, qi)(x) and cFi(qi, ri)(y), Fi(pi, ri)(x + y) < min{Fi(pi, qi)(x), Fi(qi, ri)(y)}. This is a contradiction, so (S, F) is a Menger space.
(c): Assume that (T, G) is the product of the family W based on the projection mappings {[σi] : (T, G) → (Si, Fi)}. If W doesn’t satisfy (ulp), then there is 0 < ε < 1 and {pi, qi} ⊆ Si, iI, such that inf{Fi(pi, qi)(x) | iI} ≤ δ = 1 - ε for each xR+.
Let ({*}, H) be the trivial WPMS and for each i, define the morphism [fi] : ({*}, H) → (Si, Fi) by fi(*) = pi. By assumption, there is a morphism [f] : ({*}, H) → (T, G) such that σi ° f = fi for each i.
Similarly, for each i, define the morphism [gi] : ({*}, H) → (Si, Fi) by gi(*) = qi. Then there is a morphism [g] : ({*}, H) → (T, G) such that σi ° g = gi for each i. Let s = f(*) and t = g(*). Then for each iI and xR+, Gst(x) ≤ Fi(fi(*), gi(*))(x) = Fi(pi, qi)(x), so sup(Gst) ≤ inf{Fi(pi, qi)(x)} ≤ δ, which contradicts the fact that Gst ∈ Δ+. Therefore, W satisfies (ulp).
Many authors have studied products of WPMSs independently of any categorical framework. For example, ([20], Theorem 1) establishes the finite case of Prop 4.1.1(a) and ([21], Proposition 3) “purports” to establish a general product theorem. Here are some additional details about finite products based on the previous construction:
  • If (S, F) is the product of a finite family {(Si, F(i))} of WPMS’s, then ωt = maxi,t} for each t, where ΩF = {ωt} and ΩF(i) = {ωi,t} for each i.
  • The product of a finite family of Menger spaces is a Menger space. (By Prop 2.2.2, each coordinate distance is a pseudometric, so by the previous remark, each distance in the product is a pseudometric. Hence, by Prop 2.2.2, the product is a Menger space.)
  • The product (S, F) of a finite family {(Si, Fi)} of special WPMS’s is a special WPMS. (For each p, qS and i, choose xi such that Fi(pi, qi)(xi) = 1. Then F(p, q)(max{xi}) = 1.)
The following example shows that the (ulp) property doesn’t hold in general.
Example 4.1.1.
For each nN+, define Fn : [0, n] × [0, n] → Δ+ by Fn(p, q)(x) = H(x - | p - q |).
Each ([0, n], Fn) ∈ | WPM |, but W = {([0, n] , Fn) | nN+} doesn’t satisfy (ulp). To see this, note that for each xR+, inf{Fn(0, n)(x)} = inf{H(x - n)} = 0.
By way of contrast with Prop 4.1.1, we have the following rather dramatic result.
Prop 4.1.2.
No pair of WPMSs has a coproduct in WP.
Proof 
Let (S, F) and (T, G) be WPMSs and assume that (C, H) is their coproduct in WP based on the morphisms [i] : (S, F) → (C, H) and [j] : (T, G) → (C, H). Suppose ΩH = {ρt}. Choose pS and qT and let a = i(p) and b = j(q). Let r = ρ1/2(a, b) + 1. Let D = {0, r} and let Kr0 = K0r : R+I be the characteristic function of (r, +∞). Also, let K00 = Krr = H. Then (D, K) is a WPMS and ΩK = {ωt}, where each ωt satisfies ωt(0, r) = r. By Lemma 4.1.2(b), each constant mapping defines a morphism, so there exists a morphism [h] : (C, H) → (D, K) such that h ° i = const0 and h ° j = constr. By Lemma 4.1.1, h : (C, ρ1/2) → (D, ω1/2) is non-expansive, so
ω1/2(h(a), h(b)) = ω1/2(0, r) = r ≤ρ1/2(a, b) = r - 1,
which is a contradiction. Hence, the coproduct doesn’t exist.
A result related to Prop 4.1.2 is found in ([22], Proposition 1) stating that the category Met (§4.2) has no coproducts. Next, we consider limits in WP.
Prop 4.1.3.
Suppose W = {(Si, Fi) | iD} ⊆ | WP |, where (D, ≤) is a directed set, and ({(Si, Fi) | iI}, {[fij] : (Sj, Fj) → (Si, Fi) | i, jD and ij}) is an inverse system in WP. If W satisfies (ulp), then the inverse system has an inverse limit in WP.
Proof 
Let P = ΠSi and define F : P × P → Δ+ by F(p, q) = inf{Fi(pi, qi) | iI}. By Prop 4.1.1(a), (P, F) is the product of {(Si, Fi)}. Let S = {sP | i, jD and ijfij(sj) = si}. By Lemma 4.1.2(c), (S, FS) is also a WPMS. We claim that ((S, FS), {[πi] : (S, FS) → (Si, Fi) | iD}) is the inverse limit, where each πi is a projection mapping. If ((T, G), {[φi] : (T, G) → (Si, Fi) | iI}) is a source (fij ° φj = φi for i, jD satisfying ij), then φ : TS defined by φ(t) = (φi(t)) is well-defined and πi ° φ= φi for each i. By assumption, for u, vT and iD, GuvFii(u), φi(v)), so GuvFS(φ(u), φ(v)). Therefore, [φ] : (T, G) → (S, FS) is a WP-morphism.
Prop 4.1.4.
Each directed system in WP has a direct limit.
Proof 
Let ({(Si, Fi) | iI}, {fij : (Si, Fi) → (Sj, Fj) | i, jI and ij}) be a directed system in WP, where (I, ≤) is a directed set. For each i, jI, let I(i, j) = {kI | ik and jk}. For each iI, let S^i = Si × {i}, and let S^ = ∪{S^i | iI}. For (p, i) ∈ S^i and (q, j) ∈ S^j, define (p, i) ~ (q, j) if there is kI(i, j) such that fik(p) = fjk(q). Then ~ is an equivalence relation on S^.
Let
S = {[(p, i)] : iI and pSi}
denote the set of equivalence classes for ~. Let a = [(p, i)], b = [(q, j)] ∈ S. If a = b, let Fab = H. If ab, let
Fab = sup{Fk(fik(p), fjk(q)) | kI(i, j)}.
It is easy to verify that (S, F) is a WPMS. For each iI, define ψi : SiS by ψi(p) = [(p, i)].
(1)
Each [ψi] : (Si, Fi) → (S, F) is a morphism and {[ψi]} is a sink for the directed system.
[Let a = [(p, i)], b = [(q, i)] ∈ S, and xR+. If a = b, then Fi(p, q)(x) ≤ 1 = H(x) = Fab(x). If
ab, then Fi(p, q) ≤ sup{Fk(fik(p), fik(q)) | ik} = Fab(x). If i, jI and ij, then fij(p) = fjj(fij(p)) for each pSi, so (p, i) ~ (fij(p), j). Hence, for each pSi, ψj(fij(p)) = [(fij(p), j)] = [(p, i)] = ψi(p). Therefore, ψj ° fij = ψi for each i, jI.]
(2)
((S, F), {[ψi] : (Si, Fi) → (S, F) | iI}) is the direct limit of the directed system.
[Let ((T, G), {[φi] : (Si, Fi) → (T, G) | iI}) be a sink for the directed system and define the mapping h : ST by h([(p, i)]) = φi(p). If (p, i) ~ (q, j), then there exists kI(i, j) such that fik(p) = fjk(q), so φi(p) = φk(fik(p)) = φk(fjk(q)) = φj(q). Hence, h is well-defined and h ° ψi = φi for each iI.]
We claim that [h] : (S, F) → (T, G) is a morphism. Let a = [(p, i)], b = [(q, j)] ∈ S. If a = b, then Fab = H = Gh(a)h(b). Suppose ab. If kI(i, j), then since [φk] is a morphism,
Fk(fik(p), fjk(q)) ≤ Gk(fik(p)), φk(fjk(q))) = Gi(p), φj(q)) = G(h(a), h(b)).
Therefore, FabG(h(a), h(b)).
Here are two useful examples that illustrate the previous results.
Example 4.1.2.
Each WPMS is the direct limit of the family of finite subspace WPMSs.
Let (S, F) be a WPMS and assume that I ⊆ PF+(S) satisfies S = ∪I and (I, ≤) is a directed set, where ≤ is defined by AB if AB. Let W = {(A, FA) | A ∈ I} be the family of subspace WPMSs (Lemma 4.1.2(c)) and let F = {[fAB] : (A, FA) → (B, FB) | A, B ∈ I and AB}, where each fAB is the inclusion mapping. Since {W, F} is a directed system, by Prop 4.1.4, {(T, G), {[ψA] : A ∈ I} is the direct limit, where T = {[(a, A)] : A ∈ I and aA} is the set of equivalence classes of ~ and each ψA : ST is defined by ψA(a) = [(a, A)]. Since (a, A) ~ (b, B) if there exists C ∈ I(A, B) such that fAC(a) = fBC(b), we have [(a, A)] = {(a, B) : aB ∈ I}.
By Lemma 4.1.2(c), {(S, F), {[inclA] : (S, FA) → (S, F) | A ∈ I} is a sink for the directed system, so there exists a morphism [h] : (T, G) → (S, F) such that h ° ψA = inclA for each A ∈ I. It is easy to verify that h is a bijection since S = ∪I. Also, Gtt′ = Fh(a)h(b) for each t = [(a, A)] and t′ = [(b, B)], so by Corollary 4.1.1, [h] is an isomorphism. Therefore, (S, F) is the direct limit.
Example 4.1.3.
Each Menger space (S, F) is the inverse limit of special Menger spaces with a finite range.
Let V = {V ∈ PF+(S) | {0, 1} ⊆ V} and for V, W ∈ V, define VW if VW. Then (V, ≤) is an down-directed set. Statement (1) in Theorem 3.2.2 shows that for each V ∈ V, (S, FV) is a Menger space with a finite range. Also, by Lemma 3.2.1(c), each πV = [idS] : (S, F) → (S, FV) is a morphism, so W = {(S, FV) | V ∈ V} satisfies (ulp). If V, W ∈ V and VW, then FWpq(x) ≤ FVpq(x) for each p, qS and xR+. Therefore, each fVW = [idS] : (S, FW) → (S, FV) is a morphism, so (W, {fVW | V, W ∈ V and VW}) is an inverse system in WP. Based on the construction in Prop 4.1.3, the inverse limit is (ρ, G), where ∆ = {δ ∈ SV | V, W ∈ V and VWfVWW) = δV} and G(δ, ξ) = inf{FVV, ξV) | V ∈ V} for δ, ξ ∈ ∆. Let δ ∈ ∆ and let A, B ∈ V. Let V = AB. Then V ∈ V and δA = fVAA) = δV = fVBB) = δB, so ∆ is the diagonal in SV, that is,
∆ = {δ ∈ SV | δV = δW for each V, W ∈ V}.
Since {(S, F), {πV | V ∈ V}) is a source for the inverse system, the bijection φ: S → ∆ defined by φ(s)V = s for each V ∈ V defines a morphism [φ] : (S, F) → (∆, G). Let p, qS and xR+. Then y = Fpq(x) ≤ z = Gφ(p)φ(q)(x) = inf{FVpq(x) | V ∈ V}. Choose A ∈ V that contains y. Then zFApq(x) = y, so Fpq = Gφ(p)φ(q). Hence, by Corollary 4.1.1, [φ] is an isomorphism.
Notice that the same conclusion holds by using the family V = {{0, 1}} ∪ {{0, r, 1} : 0 < r < 1}.

4.2. Subcategories of 𝓦𝓟

In this section, we introduce and compare several special subcategories of 𝓦𝓟. Here are the categories that we’ll be using:
𝓓𝓲𝓼𝓽 – objects are distance-spaces and morphisms are non-expansive mappings.
𝓟𝓜𝓮𝓽 (resp. 𝓜𝓮𝓽) – objects are pseudometric (resp. metric) spaces and morphisms are non-expansive mappings.
𝓦𝓟D – full subcategory of 𝓦𝓟 based on WPMSs determined by pseudometric spaces.
𝓦𝓟M – full subcategory of 𝓦𝓟 based on Menger spaces.
𝓦𝓟S – full subcategory of 𝓦𝓟 based on special WPMSs.
The following result describes some relationships between the various categories.
Theorem 4.2.1.
(a)
𝓟𝓜𝓮𝓽 is a reflective subcategory of 𝓓𝓲𝓼𝓽.
(b)
𝓟𝓜𝓮𝓽 and 𝓦𝓟𝓓 are isomorphic categories.
(c)
𝓦𝓟𝓓 is a coreflective subcategory of 𝓦𝓟S and the coreflection of a special WPMS (S, F) is the WPMS determined by the pseudometric space (S, ω1).
(d)
𝓦𝓟𝓜 is a reflective subcategory of 𝓦𝓟.
Proof 
(a): Let (M, ω) be a distance-space. By Prop A.2.1, there is a largest pseudometric dω such that dω ≤ ω, so r = idM : (M, ω) → (M, dω) is a non-expansive mapping. Let f : (M, ω) → (N, σ) be a non-expansive mapping to a pseudometric space and define τ by τ(x, y) = σ(f(x), f(y)). Since σis a pseudometric, τ is a pseudometric on M satisfying τ ≤ ω, so τ ≤ dω. Hence, f : (M, dω) → (N, σ) is a morphism and f ° r = f. Clearly, f is the unique mapping with this property.
(b): For each (S, γ) ∈ | 𝓟𝓜𝓮𝓽 |, let Preprints 175400 i001((S, γ)) = (S, Gγ) (Example 2.1.1) and for each non-expansive mapping g : (S, γ) → (S′, γ′), where (S′, γ′) ∈ | 𝓟𝓜𝓮𝓽 |, let Preprints 175400 i001(g) = g. By definition, for p, qS, Gpq(x) = H(x - γ(p, q)). Hence, if Gpq(x) = 1, then x > γ(p, q)) ≥ γ′(g(p), g(q)), so we obtain Gg(p)g(q)(x) = H(x - γ′(g(p), g(q))) = 1. Therefore, GpqGg(p)g(q), so Preprints 175400 i001(g) is a morphism. This defines a functor Preprints 175400 i001 : 𝓟𝓜𝓮𝓽 → 𝓦𝓟D.
For each (S, Gγ) ∈ | 𝓦𝓟D |, let 𝓕((S, Gγ)) = (S, γ) and for each morphism [h] : (S, Gγ) → (S′, Gγ′), where (S′, γ′) ∈ | 𝓦𝓟D |, let 𝓕(h) = h. Given p, qS and ε > 0, choose γ(p, q) < x < γ(p, q) + ε. Then Gpq(x) = 1, so Gh(p)h(q)(x) = 1 since [h] is a morphism. Therefore, x > γ′(h(p), h(q)). Hence, γ(p, q) + ε > γ′(h(p), h(q)) for each ε > 0, so h : (S, γ) → (S′, γ′) is a non-expansive mapping. This defines a functor 𝓕 : 𝓦𝓟D → 𝓟𝓜𝓮𝓽. Based on the definitions, Preprints 175400 i001 °𝓕 (resp. 𝓕 ° Preprints 175400 i001) is the identity functor on 𝓦𝓟D (resp. 𝓟𝓜𝓮𝓽), so Preprints 175400 i001 is an isomorphism.
(c): If (S, F) is a special WPMS, then by Lemma 2.2.2, γ = ω1 is a pseudometric. Let (S, Gγ) be the WPMS determined by (S, γ) which belongs to | 𝓦𝓟D |.
(1)
[idS] : (S, Gγ) → (S, F) is a morphism.
[Let p, qS. Since Gpq(x) = H(x - ω1(p, q)), if x ≤ ω1(p, q), then Gpq(x) = 0 ≤ Fpq(x). On the other hand, if x > ω1(p, q), then Fpq(x) = 1, so Gpq(x) ≤ Fpq(x).]
(2)
(S, Gγ) is the coreflection of (S, F) in 𝓦𝓟D.
[Suppose [g] : (M, Gγ′ = {Gab})) → (S, F) is a 𝓦𝓟-morphism where the domain WPMS is determined by the pseudometric space (M, γ′). Let a, bM. Then GabFpq, where p = g(a) and q = g(b). If γ′(a, b) < γ(p, q), choose γ′(a, b) < x < γ(p, q). Since x < γ(p, q) = ω1(p, q), Fpq(x) < 1, but Gab(x) = H(x - γ′(a, b)) = 1, so Fpq(x) = 1, which is a contradiction. Hence, γ(p, q) ≤ γ′(a, b), so g is a non-expansive mapping. Therefore, by Lemma 4.1.1, [g] : (M, Gγ′) → (S, Gγ) is a morphism, which establishes (2).]
(d): Let (S, F) be a WPMS and ΩF = {ωt}. By Prop A.2.1, for each 0 < t < 1, the path-metric ρt based on ωt is the largest pseudometric on S satisfying ρt ≤ ωt. Since ΩF is linearly ordered, Ω = {ρt} is also linearly ordered. For each 0 < t < 1, let ρt′ = sups | 0 < s < t}. By Lemma 2.4.1, {ρt′} is a linearly ordered family of pseudometrics that satisfies (lcp), so by Theorem 3.2.1(a)(c), there is a Menger space (S, G) such that ΩG = {ρt′}. Since each ρt′ ≤ ωt, by Lemma 4.1.1, [idS] : (S, F) → (S, G) is a morphism.
(3)
(S, G) is the reflection of (S, F) in 𝓦𝓟M.
[Let [f] : (S, F) → (S*, F*) be a 𝓦𝓟-morphism to a Menger space, where ΩF* = {ωt*}. By Lemma 4.1.1, for each t, f : (S, ωt) → (S*, ωt*) is non-expansive, so ωt*(f(p), f(q)) ≤ ωt(p, q) for p, qS. Since (S*, F*) is a Menger space, by Prop 2.2.2, ΩF* consists of pseudometrics, so for each t, the equation σt(p, q) = ωt*(f(p), f(q)) defines a pseudometric σt on S satisfying σt ≤ ωt. Hence, by Prop A.2.1, each σt ≤ ρt. By Prop 2.4.1(a), σt = sups | 0 < s < t} = sups | 0 < s < t} = ρt′. Hence, by Lemma 4.1.1, [f] : (S, G) → (S*, F*) is a morphism. Therefore, [idS] : (S, F) → (S, G) is the reflection mapping associated with 𝓦𝓟M.]
The following example illustrates the reflection into Menger spaces.
Example 4.2.1.
(1)
In Example 2.1.4, (R, F) is not a Menger space and the distances have the following form: for each 0 < t ≤ 1 and p, qR,
⎧ 0 | p - q | ≤ 1 - t.
ωt(p, q) = ⧼ 
⎩ 1 otherwise.
Let 0 < t < 1 and ε= 1 - t, and let ρ be a pseudometric satisfying ρ ≤ ωt. If | p - q | ≤ ε, then ρ(p, q) ≤ ωt(p, q) = 0, soρ (p, q) = 0. If | p - q | > ε, choose n satisfying 1/n < ε/| p - q | and let xk = (1 - k/n)p + (k/n)q for 0 ≤ kn. For each k, | xk - xk+1 | = | p - q |/n < ε, so ωt(xk, xk+1) = 0. Hence, ρ(p, q) ≤ Σρ(xk, xk+1) ≤ Σωt(xk, xk+1) = 0. Therefore, ρ is the zero pseudometric.
The construction used in the proof of Theorem 4.2.1(d) shows that the Menger space reflection of (R, F) is the trivial WPMS (R, {H}).
(1)
In Example 2.1.5, (R+, F) is not a Menger space and the distances have the following form: for each 0 < t ≤ 1 and p, qR+,
e1(p, q) 0 < t ≤ 1/2
ωt(p, q) = ⧼ 
e2(p, q) 1/2 < t ≤ 1.
By definition, e1(p, q) = min{a, b} and e2(p, q) = max{a, b}, where a = | p - q | and b = | p2 - q2 |. In addition, e2 is a pseudometric. Let c = 1/2, A = [0, c], and B = (c, +∞). An analysis of cases shows that the largest pseudometric d1e1 is defined by
e1(p, q) p, qA or p, qB
d1(p, q) = ⧼  q - p2 - c2 pA and qB
p - q2 - c2 qA and pB.
The construction used in the proof of Theorem 4.2.1(d) shows that the Menger space reflection (R+, G) of (R+, F) has the pseudometrics defined by
d1 0 < t ≤ 1/2
ρt′ = ⧼ 
e2 1/2 < t ≤ 1.
The construction is the one used in the proof of Theorem 3.2.1. One can show that if p, qA or p, qB, then Gpq = Fpq. However, if pA and qB, then
⎧ 0 0 ≤ xd1(p, q)
Gpq(x) = ⧼  1/2 d1(p, q) < xe2(p, q)
⎩ 1 x > e2(p, q).
For example, if p = 1/4 and q = 1, then a = 3/4 and b = 15/16. Therefore, Fpq(x) = 0 for 0 ≤ x ≤ 3/4 and Gpq(x) = 1/2 for 11/16 < x < 3/4, so FpqGpq.

5. Discussion

In some sense, the paper represents a step backwards instead of forwards since it addresses the fundamental issues rather than more contemporary work. However, the results speak for themselves. The most significant ones are (i) the construction of WPMSs from linearly ordered families (Theorems 2.4.1 and Theorem 3.2.1), (ii) the introduction of finite range WPMSs and related results (Theorems 2.3.1 and 3.2.2, and Example 4.1.3), (iii) the definition and characterization of generalized Menger spaces (Theorem 3.4.1), and (iv) the presentation of a categorical framework where the Menger spaces are a reflective subcategory (Theorem 4.2.1).
In the text, I’ve alluded to the general problem of classifying WPMSs. The originator of the subject was aware of this issue and proposed some interesting ideas in [23] that haven’t gained much attention. The geometric properties of distances that we’ve mentioned may play a role. In my opinion, the classification problem is the most interesting foundational issue. A potential sequel to the present paper may address it in some detail.

Appendix A. Distance Spaces

In this appendix, we discuss distance-spaces and present some accompanying geometric conditions that can be imposed on them.

A.1. Distances

We say that a mapping d : M × MR+ is a distance on a set M if d(x, x) = 0 for each xM and d(x, y) = d(y, x) for each x, yM. If d(x, y) > 0 for each distinct pair x, yM, then d is called a semi-metric. We say that (M, d) is a distance-space (resp. semi-metric space) if d is a distance (resp. semi-metric) on M. Let f : (M, d) → (N, e) be a mapping between distance-spaces. We say that f is uniformly continuous if for each ε > 0, there exists δ > 0 such that for each x, yM, d(x, y) < δ ⇒ e(f(x), f(y)) < ε; non-expansive if for each x, yM, e(f(x), f(y)) ≤ d(x, y), and a uniform equivalence if f is a bijection and f and f-1 are uniformly continuous.

A.2. Covers and Pseudometrics

Let 𝓒X denote the family of covers on a set X. For 𝓤, 𝓥 ∈ 𝓒X, we write 𝓤 < 𝓥 if for each U ∈ 𝓤, there is V ∈ 𝓥 such that UV. For 𝓤 ∈ 𝓒X,
𝓤δ = {St(x, 𝓤) | xX}, where St(x, 𝓤) = ∪{U ∈ 𝓤 | xU}
𝓤* = {St(V, 𝓤) | V ∈ 𝓤}, where St(V, 𝓤) = ∪{U ∈ 𝓤 | UV ≠ ∅}.
We write 𝓤 <δ 𝓥 if 𝓤δ < 𝓥 and 𝓤 <* 𝓥 if 𝓤* < 𝓥. If 𝓤, 𝓥, 𝓦 ∈ 𝓒X, 𝓤 <δ 𝓥, and 𝓥 <δ 𝓦, then 𝓤 <* 𝓦. A family {𝓤k | kN} ⊆ 𝓒X is a normal sequence if 𝓤k+1 <* 𝓤k for each k.
For a distance space (X, d) and r > 0, S(d, r) = 𝓢d(r) = {Sd(x, r) | xX} is the cover by open spheres and 𝓑d(r) = {Bd(x, r) | xX} is the cover by closed spheres, where Sd(x, r) = {yX | d(x, y) < r} and Bd(x, r) = {yX | d(x, y) ≤ r}.
For an undirected graph G = (V, E) and mapping ω : ER+, we say that P = (x0, …, xn) is a path in E from x to y if x = x0, y = xn, and each (xk, xk+1) ∈ E. Then the path-length ωP is defined by
ωP(x, y) = Σ{ω(xk, xk+1) | 0 ≤ k < n}.
The path-metric dω on V is defined by
dω(x, y) = infP | P is a path in E from x to y}
with the convention that dω(x, x) = 0 for each xV.
Prop A.2.1.
If G = (V, E) is a connected undirected graph and ω : ER+ is a mapping, then dω is the largest pseudometric on V such that dω(x, y) ≤ ω(x, y) for each (x, y) ∈ E.
The proof is left to the reader. If (M, ω) is a distance-space and G is the complete graph on M, then dω is the largest pseudometric on M satisfying dω ≤ ω.
The next result states the well-known connection between pseudometrics and normal sequences of covers.
Prop A.2.2.
Let {𝓤k} be a normal sequence of covers on a set X with 𝓤0 = {X} and let rk = 2-k for each k ≥ 0. For each x, yX, define ω(x, y) = inf{rk | ySt(x,𝓤k)}. Then (i) 𝓤k < 𝓑d(2-k) and Sd(2-k) < 𝓤kδ for each k > 0 and (ii) d ≤ ω ≤ 2d, where d = dω.
Proof 
The result (i) is found in ([24], I.14). To prove (ii), we use a statement similar to the one on ([24], page 8, line 22): (#) ≡ d(x, y) < rkySt(x, 𝓤k+1*). By (#), if d(x, y) = 0, then ySt(x, 𝓤k) for each k, so ω(x, y) = 0. If d(x, y) > 0, then either (a) d(x, y) = rk or (b) rk+1 < d(x, y) < rk for some k. (a): If k = 0, then ω(x, y) ≤ r0 = d(x, y); otherwise, by (#), d(x, y) < rk-1ySt(x, 𝓤k-1). Hence, ω(x, y) ≤ rk-1 = 2rk = 2d(x, y). (b): By (#), ySt(x, 𝓤k), so ω(x, y) ≤ rk = 2rk+1 < 2d(x, y).

A.3. Geometric Conditions

Here are two examples of geometric conditions that can be imposed on a distance-space (M, ω):
  • weak polygonal condition
    (wpc) ≡ ∀ε > 0 ∃δ > 0 ∀paths P between x, yM ● ωP < δ ⇒ ω(x, y) < ε
  • weak triangle condition
    (wtc) ≡ ∀ε > 0 ∃δ > 0 ∀x, y, zM ● ω(x, z) < δ and ω(z, y) < δ ⇒ ω(x, y) < ε.
Clearly, (wpc) implies (wtc), but the converse is false. Corresponding to each condition, there is a theorem that states a uniform equivalence with a pseudometric space. First, we establish the following result.
Prop A.3.1.
A distance-space (M, ω) satisfies (wpc) if and only if idM : (M, ω) → (M, dω) is a uniform equivalence.
Proof 
Suppose (M, ω) satisfies (wpc). Given ε > 0, there is δ > 0 such that for any path P from x to y, ωP < r ⇒ ω(x, y) < ε. If dω(x, y) < δ, then there is a path P from x to y with ωP < δ. Therefore, ω(x, y) < ε, so idM : (M, dω) → (M, ω) is uniformly continuous. Also, since dω ≤ ω, the mapping idM : (M, ω) → (M, dω) is non-expansive, so idM is a uniform equivalence. Conversely, suppose idM : (M, dω) → (M, ω) is uniformly continuous. Given ε > 0, there is δ > 0 such that for x, yM, dω(x, y) < δ ⇒ ω(x, y) < ε. If P is a path between x and y such that ωP < δ, then dω(x, y) ≤ ωP < δ. Hence, ω(x, y) < ε, so ω satisfies (wpc).
For the next result, we need to introduce the following notion. We say that a distance-space (M, ω) is a regular écart if there is a non-decreasing mapping φ: [0, +∞) → (0, +∞) that satisfies
(ec1)   For each x, y, zM, ω(x, z) < r and ω(y, z) < r ⇒ ω(x, y) < φ(r)
(ec2)   φ(0+) = φ(0) = 0.
Theorem A.3.1.
The following statements are equivalent for a distance-space (M, ω).
(a)
(M, ω) satisfies (wtc).
(b)
(M, ω) is a regular écart.
(c)
There is a path-metric d on M such that idM : (M, ω) → (M, d) is a uniform equivalence.
(d)
There is a pseudometric d on M such that idM : (M, ω) → (M, d) is a uniform equivalence.
Proof 
Given ε > 0 and r > 0, define the predicate
Pω(ε, r) ≡ ∀x, yM ● ω(x, y) ≥ ε ⇒ ∀zM ● ω(x, z) + ω(y, z) ≥ r.
Clearly, (c) ⇒ (d).
(d) ⇒ (a): Let ε > 0 and choose δ > 0 such that d(x, y) < δ ⇒ ω(x, y) < ε. Also, choose r > 0 such that ω(x, y) < rd(x, y) < δ/2. Suppose ω(x, y) ≥ ε and let zM. Then d(x, y) ≥ δ, so without loss of generality, we can assume that d(x, z) ≥ δ/2. Therefore, ω(x, z) + ω(y, z) ≥ ω(x, z) ≥ r. Hence, Pω(ε, r) holds, so (M, ω) satisfies (wtc).
To make the remaining proof less cluttered, we put the proofs of statements (1) – (4) at the end.
(b) ⇒ (c): By assumption, there is a non-decreasing mapping φ : (0, +∞) → (0, +∞) that satisfies (ec1) and (ec2), so we can assume that φ(0) = 0. Then the following statements hold.
(1)
For each xM and r > 0, St(x, Sω(r)) ⊆ Sω(x, φ(r)).
(2)
There is a decreasing sequence {εk} ⊆ (0, 1] converging to 0 with ε0 = ε1 = 1 such that φ(εk+1) < εk and Sωk+1) <* Sωk) for each k > 0.
Let {εk} be a sequence that satisfies the conditions in (2). Let 𝓪0 = {M} and 𝓪k = Sωk) for k > 0. By (2), {𝓪k} is a normal sequence. Define the distance σ on M by σ(x, y) = inf{2-k | ySt(x, 𝓪k)} and let dσ be the induced path-metric. Then by Prop A.2.2(ii), dσ ≤ σ ≤ 2dσ, so σ and dσ are uniformly equivalent.
(3)
ω and σ are uniformly equivalent, so dσ and ω are uniformly equivalent. Hence, part (c) holds.
(a) ⇒ (b): Without loss of generality, assume that Δ = diamω(M) > 0.
Case 1: Δ = +∞ or Δ < +∞ and Rng(ω) ⊆ [0, Δ).
(4)
(i): If Δ = +∞, then there is a non-decreasing g : (0, +∞) → (0, +∞) such that Pω(ε, g(ε)) holds for each ε > 0. (ii) If Δ < +∞ and Rng(ω) ⊆ [0, Δ), then there is a non-decreasing g : (0, Δ) → (0, Δ] such that Pω(ε, g(ε)) holds for each 0 < ε < Δ.
Subcase 1: L = limε → Δ-g(ε) = +∞.
For each r > 0, Br = {ε > 0 | g(ε) ≥ r} ≠ ∅. Define τ : (0, +∞) → [0, +∞) by τ(r) = inf(Br). If 0 < r < r′, then Br′Br, so τ(r) ≤ τ(r′). Hence, τ is non-decreasing. Define φ : (0, +∞) → (0, +∞) by φ(r) = τ(2r) + r. Then φ is also non-decreasing.
Suppose x, yM and ω(x, y) ≥ φ(r) for some r > 0. Let zM and let s = ω(x, z) + ω(y, z). Since ω(x, y) > τ(2r) = inf(B2r), there is ε ∈ B2r such that ω(x, y) > ε. Then by (4), sg(ε) ≥ 2r, so either ω(x, z) ≥ r or ω(y, z) ≥ r. Hence, φ satisfies (ec1).
Subcase 2: L = limε → Δ-g(ε) < +∞.
For each 0 < r < L, Br = {0 < ε < Δ | g(ε) ≥ r} ≠ ∅. Define τ : (0, L) → [0, Δ) by τ(r) = inf(Br). As in Subcase 1, τ is non-decreasing, so K = limr → L-τ(r) exists and K ≤ Δ. Now define τ on (0, +∞) by assigning τ(r) = K for each rL. Then τ : (0, +∞) → (0, Δ] is non-decreasing. Define φ : (0, +∞) → (0, +∞) by φ(r) = τ(2r) + 4rΔ/L. Then φ is also non-decreasing.
Suppose x, yM and ω(x, y) ≥ φ(r) for some r > 0. Let zM and let s = ω(x, z) + ω(y, z). If rL/2, then ω(x, y) ≥ 4rΔ/L ≥ 2Δ, which is impossible. Since r < L/2, ω(x, y) > τ(2r) = inf(B2r), so there is ε ∈ B2r such that ω(x, y) > ε. As in Subcase 1, by (4), either ω(x, z) ≥ r or ω(y, z) ≥ r, so φsatisfies (ec1).
If a = limn → +∞τ(1/n) > 0, choose 0 < x < a. If L < +∞, then for each n > 1/L, x < τ(1/n) = inf(B1/n), so g(x) < 1/n. Hence, g(x) = 0, which is a contradiction. Hence, a = 0, so φ(0+) = 0. If L = +∞, the same proof works, so in either case, φsatisfies (ec2).
Case 2: Δ < +∞ and Δ ∈ Rng(ω).
Let M′ = Md B, where B = (0, Δ + 1) and define ω′ : M′ × M′ → R+ by ω′(x, y) = ω(x, y) if x, yM, ω′(x, y) = | x - y | if x, yB, and ω′(x, y) = ω′(y, x) = 1 if xM and yB. Evidently, diamω′(M′) = Δ + 1 and Rng(ω′) ⊆ [0, Δ + 1). By part (a), given ε > 0, choose r > 0 such that Pω(ε, r) holds. Then Pω′(ε, min{r, ε, 1}) holds, so (M′, ω′) satisfies (wtc). Hence, by Case 1, (M′, ω′) is a regular écart, so (M, ω) is also a regular écart.
Proofs (intermediate statements)
(1)
For each xM and r > 0, St(x, Sω(r)) ⊆ Sω(x, φ(r)).
Let xM. If xSω(p, r) for some pM, then ω(x, p) < r and ω(p, y) < r for each ySω(p, r), so by (ec1), ω(x, y) < φ(r). Hence, St(x, Sω(r)) ⊆ Sω(x, φ(r)).
(2)
There is a decreasing sequence {εk} ⊆ (0, 1] converging to 0 with ε0 = ε1 = 1 such that φ(εk+1) < εk and Sωk+1) <* Sωk) for each k > 0.
Let ε0 = ε1 = 1 and suppose ε1, …, εk have been defined for k > 0 such that the two conditions hold. By (ec2), there is r > 0 such that r < εk and φ(r) < εk and there is 0 < εk+1 < εk such that φ(εk+1) < r. Then φ(εk+1) < εk and by (1), Sωk+1) <δ Sω(φ(εk+1)) and Sω(φ(εk+1)) < Sω(r) <δ Sω(φ(r)) < Sωk). Hence, by our earlier remarks, Sωk+1) <* Sωk).
(3)
ω and σ are uniformly equivalent, so dσ and ω are uniformly equivalent.
Let x, yM. If σ(x, y) ≤ 2-(k+1) for some k ≥ 0, then ySt(x, 𝓪k+1), so there is zM such that x, ySω(z, εk+1). Hence, by (ec1) and (2), ω(x, y) < φ(εk+1) < εk. Conversely, if ω(x, y) < εk for some k > 1, then ySt(x, 𝓪k), so σ(x, y) ≤ 2-k.
(4)
(i): If Δ = +∞, then there is a non-decreasing g : (0, +∞) → (0, +∞) such that Pω(ε, g(ε)) holds for each ε > 0. (ii) If Δ < +∞ and Rng(ω) ⊆ [0, Δ), then there is a non-decreasing g : (0, Δ) → (0, Δ] such that Pω(ε, g(ε)) holds for each 0 < ε < Δ.
By part (a), for each ε > 0, Sε = {r > 0 | Pω(ε, r)} ≠ ∅.
(i): Let ε > 0 and choose x, yM such that ω(x, y) > ε. If rSε, then ω(x, y) + ω(y, y) ≥ r, so Sε ⊆ (0, ω(x, y)]. Define g : (0, +∞) → (0, +∞) by g(ε) = sup(Sε). If 0 < ε < ε′, then SεSε′, so g(ε) ≤ g(ε′). Hence, g is non-decreasing. If x, yM satisfies ω(x, y) ≥ ε > 0, then for each rSε and zM, ω(x, z) + ω(y, z) ≥ r. Hence, ω(x, z) + ω(y, z) ≥ g(ε), so Pω(ε, g(ε)) holds.
(ii): Let 0 < ε < Δ. The argument used in (i) shows that Sε ⊆ (0, Δ]. Define g : (0, Δ) → (0, Δ] by g(ε) = sup(Sε). Then the previous argument shows that each Pω(ε, g(ε)) holds.
In [17], the (wtc) property is called Axiom V and the definition of a regular écart is attributed to Fréchet. A regular écart is also referred to as property (V) in [14] and [15]. We can call Theorem A.3.1 the Chittenden-Wilson Theorem since the equivalence of (a) and (b) is established in ([17], Theorem I), (b) ⇒ (d) is proved in [15], and (d) ⇒ (b) is proved in ([17], Theorem II). Our proof that (a) ⇒ (b) is a modified and more detailed version of the original proof and our proof that (b) ⇒ (c) differs significantly from the original proof.

Funding

This research received no external funding.

Conflicts of Interest

The author declares no conflict of interest.

References

  1. B. Schweizer and A. Sklar, Probabilistic Metric Spaces, North-Holland, 1983. MR790314 (no DOI).
  2. E. P. Klement, R. Mesiar, and E. Pap, Triangular Norms, Kluwer, Dordrecht, 2000. MR1790096. [CrossRef]
  3. M. Liu, A representation theorem for probabilistic metric spaces in general, Czech. J. Math. 50, 3(2000), 551-554. MR1777476. [CrossRef]
  4. D. Hofmann and C. D. Reis, Probabilistic metric spaces as enriched categories, Fuzzy Sets and Systems 210 (2013), 1-21. MR2981702. [CrossRef]
  5. E. Nishiura, Constructive methods in probabilistic metric spaces, Fund. Math. 67 (1970), 115-124. MR259978. [CrossRef]
  6. M. Bachir and B. Nazaret, Metrization of probabilistic metric spaces. Applications to fixed point theory and Arzela-Ascoli type theorem, Topology and its Applications 289 (2021). MR4195115. [CrossRef]
  7. V. H. Badshah, S. Jain, and S. Mandloi, Fixed Point Theorem and Semi-Compatibility in Menger Probabilistic Metric Space, Annals of Pure and Applied Mathematics 14, 3(2017), 407-415. [CrossRef]
  8. S. Chang, Y. J. Cho, and S. Kang, Nonlinear Operator Theory in Probablistic Metric Spaces, Nova Publishers, 2001. MR2018691 (no DOI).
  9. Y. Kurihara, Stochastic metric space and quantum mechanics, arXiv:1612.04228 v5, 2017. [CrossRef]
  10. V. Torra and G. Navarro-Arribas, G. (2018). Probabilistic Metric Spaces for Privacy by Design Machine Learning Algorithms: Modeling Database Changes. In: Garcia-Alfaro, J., Herrera-Joancomartí, J., Livraga, G., Rios, R. (eds) Data Privacy Management, Cryptocurrencies and Blockchain Technology. DPM CBT 2018 2018. Lecture Notes in Computer Science 11025. Springer, Cham. [CrossRef]
  11. R.R. Stevens, Metrically generated probabilistic metric spaces, Fund. Math. 61 (1968), 259-269. MR250353. [CrossRef]
  12. H. Sherwood, On E-spaces and their relation to other classes of probabilistic metric spaces, J. London Math. Soc. 44 (1969), 441-448. MR0240845. [CrossRef]
  13. J. H. Li and J. X. Fang, The family of Δ-pseudometrics and Menger probabilistic metric spaces, December, 1997, ResearchGate (no DOI).
  14. M. Fréchet, Les ensembles abstraits et le calcul fonctionnel, Rendiconti del Circolo Matemático di Palermo 30 (1910), 1-26. [CrossRef]
  15. E. W. Chittenden, On the equivalence of écart and voisinage, Trans. AMS. 18 (1917), 161-166. MR1501066. [CrossRef]
  16. V. W. Niemytzki, On the third axiom of metric space, Trans. AMS. 29, 3(1927), 507-513. MR1501402. [CrossRef]
  17. W. A. Wilson, On semi-metric-spaces, Amer. J. Math. 53, 2(1931), 361-373. MR1506845. https://doi.org/10.2307/2370790.
  18. R. C. Flagg, Quantales and continuity spaces, Algebra Univers. 37 (1997) 257-276. MR1452402. https://doi.org/10.1007/s000120050018.
  19. J. Adámek, H. Herrlich, and G. Strecker, Abstract and concrete categories: the joy of cats, Repr. Theory Appl. Categ. 17 (2006), 1-507. MR2240597 (no DOI).
  20. R. Egbert, Products and quotients of probabilistic metric spaces, Pacific J. Math. 24 (1968), 437-455. MR0226690. [CrossRef]
  21. I. Goleţ, On uncountable product of probabilistic metric spaces, World Applied Sciences Journal 6, 9 (2009), 1304-1308. (no DOI).
  22. V. A. Lemin, Finite ultrametric spaces and computer science, in "Categorical Perspectives", ed. J. Koslowski, A. Melton, Trends in Mathematics, v. 16, Birkhauser Verlag, Boston - Basel - Berlin, 2001, 219-242. MR1827671. [CrossRef]
  23. K. Menger, Probabilistic geometry, Proc. NAS. U.S.A. 37 (1951), 226-229. MR0042081 (no DOI).
  24. J. R. Isbell, Uniform Spaces, Mathematical Surveys 12, Amer. Math. Society, Providence, 1964. MR0170323. [CrossRef]
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

Disclaimer

Terms of Use

Privacy Policy

Privacy Settings

© 2025 MDPI (Basel, Switzerland) unless otherwise stated