Formally Real Rings and R. E. Johnson’s Theorem

Q: Does a formally real ring have a canonical ordering?

No. The only canonical object is the weak preordering T(R), which is the intersection of nothing and is usually far from total. Selecting an ordering requires Zorn's Lemma and involves genuine choice; a ring with several orderings has no distinguished one.

Q: If R is formally real, is its quotient ring or its ring of fractions formally real?

When a ring of quotients exists in the sense of (17.17), yes: every ordering of R extends uniquely to it, so it is orderable and hence formally real. Without that hypothesis, extendability is governed by the criterion of (17.16) and can fail.

Q: Are there formally real rings that are not domains?

Not under Lam's definition — (17.7) forces a formally real ring to be a domain. Commutative real algebra uses a weaker notion, also called *real*, which permits zero divisors; that theory is about the real spectrum of a general commutative ring rather than about total orderings of the ring itself.

Executive Summary

Call a nonzero ring $R$ formally real if $0$ is not a sum of permuted products $per (a_{1}^{2} \dots a_{m}^{2})$ with all $a_{i} \neq 0$ . The set of such sums, written $T (R)$ , is the weak preordering: it satisfies both preordering axioms automatically, sits inside every preordering of $R$ , and is a preordering exactly when $R$ is formally real.

R. E. Johnson's theorem $(17.11)$ then closes the circle: formally real, admitting a preordering, and admitting an ordering are three names for the same property. Combined with $(17.7)$ this is a complete, checkable answer to the orderability problem for arbitrary rings, and it specialises to Artin and Schreier's 1927 criterion when $R$ is a field.

$T (R)$ Smallest preordering

3Equivalent conditions

1927Artin–Schreier

1952Johnson

Overview

The obstruction to ordering a ring must be finitary if the theory is to be usable: one wants a single equation whose absence guarantees an ordering. $(17.4)$ and $(17.7)$ show what such an equation must look like. In any ordering, a product in which every factor is repeated an even number of times is positive; so is any sum of such products; and $0$ is not positive. Formal reality simply asserts that this necessary condition holds.

T (R) = {textstyle \sum_{k} per (a_{k 1}^{2} \dots a_{k m_{k}}^{2}) : a_{k j} \in R ∖ {0}}, R formally real ⟺ 0 \notin T (R) .

(17.11a)

The weak preordering and the definition of formal reality. Note $1 = per (1^{2}) \in T (R)$ always, so $T (R) \neq \emptyset$ .

The proof of sufficiency is short because the work has already been done. If $0 \notin T (R)$ then $T (R)$ is a preordering; Zorn's Lemma enlarges it to a maximal preordering; and by $(17.10)$ a maximal preordering is an ordering. The substance is in Preorderings in Rings; this page is where the payment is collected.

Learning Objectives

Write down $T (R)$ and check that it satisfies $(17.5)$ and $(17.6)$ for any ring.
Prove $T (R) \subseteq T$ for every preordering $T$ of $R$ .
Prove all three implications in R. E. Johnson's theorem $(17.11)$ .
Show that a field is formally real precisely when $- 1$ is not a sum of squares.
Prove that a division ring $D$ is not formally real precisely when $- 1 \in T (D)$ .
State the four equivalent descriptions of the division closure $\bar{T}$ in $(17.12)$ .

Definitions

Definition(17.11a)Weak preordering and formal reality

For a ring $R \neq 0$ , let $T (R)$ be the set of all finite sums of elements $per (a_{1}^{2} \dots a_{m}^{2})$ with $m \geq 1$ and $a_{i} \in R ∖ {0}$ — that is, products of finitely many nonzero elements, each occurring an even number of times, multiplied in any order. $R$ is formally real if $0 \notin T (R)$ .

$T (R)$ always satisfies $(17.5)$ , being closed under sums by construction, and $(17.6)$ , since inserting doubled elements or elements of $T (R)$ into a permuted product produces another such sum. Hence: $R$ is formally real if and only if $T (R)$ is a preordering, and in that case $T (R)$ is the smallest preordering of $R$ .

$T (R)$: The weak preordering. In a commutative ring it is the set of nonzero sums of squares.
Totally positive: Positive in every ordering of $R$ . By (17.13) this is membership in the division closure $\bar{T (R)}$ , not in $T (R)$ itself.
$\bar{T}$: The division closure of a preordering $T$ : those $a$ with $a t \in T$ for some $t \in T$ .
Real field: Synonym for formally real field in much of the real-algebra literature; a real closed field is one with no proper formally real algebraic extension.

Formal reality is a property of the ring alone — no cone has been chosen. Choosing a cone is extra structure, and $(17.11)$ says the choice is possible exactly when the property holds.

Core Concepts

Why $T (R)$ is the right generating set

Any preordering must contain $per (a_{1}^{2} \dots a_{m}^{2})$ by axiom $(17.6)$ with $n = 0$ , and must be closed under sums by $(17.5)$ . Therefore it contains all of $T (R)$ . Dually, $T (R)$ imposes no further constraints. It is the free object of the theory: the preordering generated by nothing at all.

T (R)

any preordering

T

any ordering

P

The commutative shadow

If $R$ is commutative, $per (a_{1}^{2} \dots a_{m}^{2}) = (a_{1} \dots a_{m})^{2}$ , so $T (R)$ is exactly the set of nonzero sums of squares, and formal reality says that a sum of squares of nonzero elements is never $0$ . For a field this is the classical Artin–Schreier condition, as the next section makes precise.

What formal reality already implies

By $(17.7)$ applied to $T (R)$ : a formally real ring is a domain of characteristic zero in which $1$ is totally positive. So formal reality is strictly stronger than being a characteristic-zero domain — the real quaternions separate the two conditions, since $- 1 = i^{2} \in T (ℍ)$ makes $0 = 1 + i^{2} \in T (ℍ)$ .

Key Results

Theorem(17.11)R. E. Johnson

For any ring $R \neq 0$ the following are equivalent:

$R$ is formally real, i.e. $0 \notin T (R)$ ;
$R$ possesses a preordering;
$R$ possesses an ordering.

Proof

**(3) $\Rightarrow$ (2).** Every ordering is a preordering: by the sign homomorphism of an ordered ring, a product with all multiplicities even is positive, and positives are closed under addition.

**(2) $\Rightarrow$ (1).** Let $T$ be a preordering. Axiom $(17.6)$ with $n = 0$ puts every $per (a_{1}^{2} \dots a_{m}^{2})$ into $T$ , and axiom $(17.5)$ then puts every sum of such into $T$ . Hence $T (R) \subseteq T \subseteq R ∖ {0}$ , so $0 \notin T (R)$ .

**(1) $\Rightarrow$ (2).** If $0 \notin T (R)$ then $T (R) \subseteq R ∖ {0}$ , and since $T (R)$ satisfies $(17.5)$ and $(17.6)$ by construction, it is a preordering.

**(2) $\Rightarrow$ (3).** Given a preordering $T$ , Zorn's Lemma provides a maximal preordering $T_{1} \supseteq T$ : the union of a chain of preorderings satisfies both axioms, because each axiom involves only finitely many elements, and omits $0$ because every member does. By $(17.10)$ , $T_{1}$ is an ordering.

Corollary(17.11b)Structure of a formally real ring

A formally real ring is a domain of characteristic zero; its only idempotents are $0$ and $1$ ; it has no nonzero nilpotent elements; and it is infinite. All of this follows from $(17.7)$ applied to $T (R)$ , together with $(17.11)$ .

Proposition(17.11c)The field and division ring case

Let $D$ be a division ring. Then $D$ fails to be formally real if and only if $- 1 \in T (D)$ . In particular a field $F$ is formally real if and only if $- 1$ is not a sum of squares in $F$ , which is the Artin–Schreier criterion.

Proof

If $- 1 \in T (D)$ then $0 = 1 + (- 1) \in T (D) + T (D) \subseteq T (D)$ , so $D$ is not formally real.

Conversely suppose $0 \in T (D)$ , say $0 = p_{1} + p_{2} + \dots + p_{N}$ where each $p_{k} = per (a_{k 1}^{2} \dots)$ is a permuted doubled product of nonzero elements. Each $p_{k}$ is a product of nonzero elements of a division ring, hence $p_{k} \neq 0$ ; therefore $N \geq 2$ and

- p_{1} = p_{2} + \dots + p_{N} \in T (D) .

(J.1)

Next, $p_{1}^{- 1}$ itself lies in $T (D)$ . Indeed, if $p_{1}$ is an arrangement of the multiset ${a_{1}, a_{1}, \dots, a_{m}, a_{m}}$ , then the arrangement $p_{1}^{- 1}, p_{1}^{- 1}$ followed by that same multiset is again a permuted product with all multiplicities even, and its value is

(p_{1}^{- 1})^{2} p_{1} = p_{1}^{- 1} .

(J.2)

Finally $T (D)$ is closed under multiplication, so $(- p_{1}) \cdot p_{1}^{- 1} = - 1$ lies in $T (D)$ .

For a field, $T (F)$ is the set of nonzero sums of squares, so the criterion reads: $F$ is not formally real if and only if $- 1$ is a sum of squares.

Remark

For a general ring the reduction of $0 \in T (R)$ to $- 1 \in T (R)$ is not available, since the inverse used in $(J . 2)$ need not exist; Lam poses the corresponding question for integral domains as Exercise 6 of §17. The definition via $0 \notin T (R)$ is the one that makes $(17.11)$ true, and it should be taken as primary.

Proposition(17.12)Four descriptions of the division closure

Let $T$ be a preordering in a ring $R$ . Then the following four subsets of $R$ coincide; the common value is written $\bar{T}$ and called the division closure of $T$ :

{a : a t \in T for some t \in T} = {a : t' a \in T for some t' \in T} = {a : a b^{2} \in T for some b \neq 0} = {a : b^{2} a \in T for some b \neq 0} .

(17.12)

Moreover $T \subseteq \bar{T}$ and $0 \notin \bar{T}$ .

Proof

Suppose $a t = t'$ with $t, t' \in T$ . Then $t' a = a t a \in T$ by $(17.6)$ , so $a$ lies in the second set; and $a t^{2} = (a t) t = t' t \in T$ with $t \neq 0$ , so $a$ lies in the third set with $b = t$ . The reverse inclusions are immediate: $b \neq 0$ gives $b^{2} \in T$ , so $a b^{2} \in T$ exhibits a witness $t = b^{2}$ for the first set, and likewise for the fourth.

Symmetrically, if $t' a = t'' \in T$ then $a t'' = a t' a \in T$ places $a$ in the first set, and $t'^{2} a = t' (t' a) \in T$ places it in the fourth. Chasing these four implications round gives equality of all four sets.

For $t \in T$ we have $t \cdot t \in T$ , so $T \subseteq \bar{T}$ . And $0 \notin \bar{T}$ because $0 \cdot t = 0 \notin T$ .

Proof Techniques and Method

How these proofs work, and which move to reuse.

Move 1

Generate, then test properness

Build the smallest set closed under the required operations; the only thing that can go wrong is that it contains $0$ . This turns an existence problem into the non-existence of one identity.

Move 2

Maximise by Zorn, then use maximality

Maximal objects in this theory are automatically total. The pattern — enlarge, then read off totality from $(17.10)$ — replaces any attempt to define the order element by element.

Move 3

Invert inside the cone

In a division ring, $(p^{- 1})^{2} p = p^{- 1}$ shows the cone is closed under inversion. Wherever inverses exist, positivity statements can be divided as well as multiplied.

Move 3 is exactly what is unavailable in a general ring, and it is why the division closure $\bar{T}$ has to be introduced at all: it is the smallest correction that restores the ability to divide by positives. That correction is the subject of Division-Closed Preorderings.

Worked Example

Formally real, and not

$ℚ$ is formally real: a sum of nonzero rational squares is a positive rational, never $0$ . So is $ℝ$ , and so is $ℚ (\sqrt{2})$ , which has two orderings. $ℚ (i)$ is not: $1^{2} + i^{2} = 0$ exhibits $0 \in T (ℚ (i))$ directly, with $m = 2$ and both elements nonzero.

For the quaternions, $1^{2} + i^{2} = 0$ again, so $ℍ$ is not formally real even though it is a characteristic-zero division ring. Consistently, $- 1 = i^{2} \in T (ℍ)$ as predicted by $(17.11 c)$ .

A noncommutative formally real ring

Let $A = ℝ ⟨ x, y ⟩$ be the free $ℝ$ -algebra on two generators. $A$ is orderable — order it by any total order on the free monoid compatible with multiplication, deciding positivity by the coefficient of the least word in the support — so by $(17.11)$ it is formally real. The same construction applies over any formally real coefficient field.

A calculation in $T (R)$

In $R = ℝ ⟨ x, y ⟩$ , the element $x y x y + y x^{2} y + 3$ lies in $T (R)$ : it is $per (x^{2} y^{2})$ in the arrangement $x y x y$ , plus $per (y^{2} x^{2})$ in the arrangement $y x x y$ , plus $1 + 1 + 1$ . None of these is a square in the free algebra, which is what the permuted-product notation is for.

x y x y + y x^{2} y + 3 = per (x^{2} y^{2}) + per (x^{2} y^{2}) + per (1^{2}) + per (1^{2}) + per (1^{2}) \in T (R) .

(E.1)

Two different arrangements of the same multiset ${x, x, y, y}$ , plus three copies of $1$ .

Totally positive is not the same as in $T (R)$

In $R = ℝ [x]$ , the polynomial $x^{2} + 1$ lies in $T (R)$ outright. The subtlety appears for elements $a$ with $a b^{2} \in T (R)$ but $a \notin T (R)$ : such an $a$ is positive in every ordering yet is not itself a sum of squares. Lam constructs an explicit commutative example in Exercise 8 of §17; Artin proved in Exercise 9 that no such example exists in $k [t]$ for $k$ a formally real field.

Comparison and Classification

Formal reality across standard rings
Ring	Formally real?	Witness
$ℚ$ , $ℝ$ , $ℚ (\sqrt{2})$	yes	sums of nonzero squares are positive reals
$ℚ (i)$ , $ℂ$	no	$1^{2} + i^{2} = 0$
$ℍ$	no	$1^{2} + i^{2} = 0$ ; a division ring nonetheless
$𝔽_{p}$ , $ℤ / n ℤ$	no	characteristic is not $0$
$M_{2} (ℝ)$	no	$e_{12}$ gives $per (e_{12}^{2}) = 0$
$ℝ [x]$ , $ℝ (x)$	yes	orderable by leading coefficient
$ℝ ⟨ x, y ⟩$	yes	ordered via an ordered free monoid
$A_{1} (ℝ)$ (Weyl algebra)	yes	ordered by top $y$ -coefficient
$ℝ [x] / (x^{2} + 1)$	no	isomorphic to $ℂ$

Which conditions imply which
	Domain	Char $0$	Formally real	Orderable
Formally real	yes	yes	yes	yes
Orderable	yes	yes	yes	yes
Char $0$ domain	yes	yes	no	no
Domain	yes	no	no	no

Which conditions imply which

Reading a row gives the properties it implies. The two lower rows fail to imply formal reality: $ℍ$ is the standard counterexample.

Relationship Map

$R$ formally real — $0 \notin T (R)$
- equivalent to
  - $R$ has a preordering, $(17.11)$
  - $R$ has an ordering, $(17.11)$
- implies
  - $R$ is a domain, $char R = 0$ , $(17.7)$
  - $1$ is totally positive
  - $R$ is infinite, with no nonzero nilpotents
- does not follow from
  - being a domain of characteristic $0$ — see $ℍ$
  - being a division ring
- specialises to
  - fields: $- 1$ is not a sum of squares (Artin–Schreier)
  - division rings: $- 1 \notin T (D)$ , $(17.11 c)$

formally real

T (R)

is a preorderingmaximal preordering existsordering exists

Applications and Industry Use

Applications here means where this structure is used — inside mathematics and in the engineering and computing disciplines that consume it.

Real algebra

Real closed fields and model theory

Formally real fields have real closures, and the first-order theory of real closed fields admits quantifier elimination. That single fact underwrites every algorithmic decision procedure for polynomial inequalities.

Quantum foundations

Formally real Jordan algebras

The Jordan–von Neumann–Wigner classification of finite-dimensional formally real Jordan algebras — where a sum of squares vanishes only if each term does — was motivated by the search for algebraic models of observables in quantum mechanics.

Optimisation

Sums-of-squares relaxations

For commutative polynomial rings, membership in $T (R)$ up to a degree bound is a semidefinite feasibility problem. This is the computational core of polynomial optimisation and of Lyapunov certificate search in control.

Group rings

Zero divisors and units

If $G$ is an orderable group then $k G$ is a domain for any domain $k$ , by the least-support-term argument. Formal reality of coefficient rings then propagates orderability, giving large families of orderable noncommutative rings.

Inside ring theory itself, the practical value of $(17.11)$ is that orderability becomes a closure condition on generators rather than a construction: to prove a ring orderable, it suffices to rule out one kind of identity.

Standards and Notation

Standards here covers notation, symbol and markup standards, and reference implementations, rather than material or design codes.

Weak preordering

T (R)

in Lam;

\sum R^{2}

in commutative real algebra

Formally realalso real (Prestel, Bochnak–Coste–Roy), and semireal for a weaker variant in some sources

Division closure

\bar{T}

; sometimes written

T^{sat}

or the saturation of

T

Real spectrum

Sper R

, the space of orderings of a commutative ring

ImplementationsQEPCAD B, Redlog and SMT solvers with nonlinear real arithmetic decide real feasibility; SOSTOOLS and SumsOfSquares.jl certify membership in cones

MarkupMathML per ISO/IEC 40314; ordering relations per ISO 80000-2

Failure Modes and Common Mistakes

Do not conclude formal reality from the absence of zero divisors; $ℍ$ refutes that.
Do not assume a subring of a formally real ring inherits anything more than formal reality — the number of orderings can change drastically on passing to a subring or an extension.
Do not forget that $T (R)$ is defined with $a_{i} \neq 0$ ; allowing $a_{i} = 0$ would put $0$ in $T (R)$ trivially and make every ring non-formally-real.
Do not expect formal reality to be preserved by quotients. $ℝ [x]$ is formally real but $ℝ [x] / (x^{2} + 1) ≅ ℂ$ is not.

Historical Notes and Lessons Learned

1927Artin and SchreierFormally real fields are defined by the condition that $- 1$ is not a sum of squares, and shown to be exactly the orderable fields. Real closures are constructed and shown to be unique.
1927Artin and Hilbert's 17th problemA rational function over $ℝ$ that is nonnegative wherever defined is a sum of squares of rational functions — the field-theoretic prototype of the totally-positive criterion $(17.15)$ .
1934Formally real Jordan algebrasJordan, von Neumann and Wigner classify finite-dimensional formally real Jordan algebras, transferring the notion out of associative algebra entirely.
1940sOrdered noncommutative ringsAlbert, B. H. Neumann and Fuchs develop orderings on noncommutative rings, quotient rings and division rings built from ordered groups.
1952R. E. JohnsonThe Artin–Schreier equivalence is proved for arbitrary rings once the correct notion of formal reality — no vanishing sum of permuted doubled products — is identified.
1960s–70sSerre's preorderingsThe preordering formulation streamlines the proof and becomes the standard route; it also transplants directly to the theory of quadratic forms and to the real spectrum.

The lesson: the theorem was blocked for twenty-five years not by a missing argument but by a missing definition. Once formal reality was expressed in terms of permuted products rather than squares, the classical proof went through with no new ideas.

Quick Reference

Weak preordering

T (R) =

sums of

per (a_{1}^{2} \dots a_{m}^{2})

, all

a_{i} \neq 0

Formally real

0 \notin T (R)

(17.11)formally real

\Leftrightarrow

has a preordering

\Leftrightarrow

has an ordering

Minimality

T (R) \subseteq T

for every preordering

T

Consequencesdomain,

char R = 0

1

totally positive

Fieldsformally real

\Leftrightarrow

- 1

is not a sum of squares

Division ringsnot formally real

\Leftrightarrow

- 1 \in T (D)

Division closure

\bar{T} = {a : a b^{2} \in T

for some

b \neq 0}

Certificates in both directions
To prove	Exhibit
$R$ is not formally real	one identity $0 = \sum_{k} per (\dots)$ with all factors nonzero
$R$ is formally real	an explicit ordering, or an ordered ring containing $R$
$R$ has at least two orderings	an element $b$ with $T (R)_{b}$ and $T (R)_{- b}$ both proper
$a$ is totally positive	some $b \neq 0$ with $a b^{2} \in T (R)$ , by $(17.15)$

Frequently Asked Questions

Why is formal reality defined with permuted products rather than squares?

Because in a noncommutative ring the elements forced to be positive by an ordering are precisely those products in which every factor is repeated an even number of times, in any arrangement — $a b a b$ is positive, though it is not a square. Defining $T (R)$ with squares only would give a set too small to contain every preordering, and $(17.11)$ would fail.

Is a subring of a formally real ring formally real?

Yes, immediately: an ordering restricts to any subring, so by $(17.11)$ formal reality is inherited. The converse fails badly — $ℝ$ is formally real but $ℂ$ is not, and $ℝ [x]$ is formally real while many of its quotients are not.

How is this different from the Artin–Schreier theorem?

Artin–Schreier is the field case, where the criterion is that $- 1$ is not a sum of squares. $(17.11)$ keeps the shape of the statement but replaces both the notion of square and the notion of cone by their noncommutative counterparts, and works for arbitrary rings with no commutativity, finiteness or invertibility hypotheses.

Does a formally real ring have a canonical ordering?

No. The only canonical object is the weak preordering $T (R)$ , which is the intersection of nothing and is usually far from total. Selecting an ordering requires Zorn's Lemma and involves genuine choice; a ring with several orderings has no distinguished one.

R

is formally real, is its quotient ring or its ring of fractions formally real?

When a ring of quotients exists in the sense of $(17.17)$ , yes: every ordering of $R$ extends uniquely to it, so it is orderable and hence formally real. Without that hypothesis, extendability is governed by the criterion of $(17.16)$ and can fail.

Are there formally real rings that are not domains?

Not under Lam's definition — $(17.7)$ forces a formally real ring to be a domain. Commutative real algebra uses a weaker notion, also called real, which permits zero divisors; that theory is about the real spectrum of a general commutative ring rather than about total orderings of the ring itself.

References

T. Y. Lam, A First Course in Noncommutative Rings, Graduate Texts in Mathematics 131, Springer-Verlag, 1991, §17, results (17.11)–(17.12), pp. 279–281.
R. E. Johnson, On ordered domains of integrity, American Mathematical Monthly 59 (1952). The original source of the equivalence in (17.11).
E. Artin and O. Schreier, Algebraische Konstruktion reeller Körper, Abhandlungen aus dem Mathematischen Seminar der Universität Hamburg 5 (1927).
A. Prestel, Lectures on Formally Real Fields, Lecture Notes in Mathematics 1093, Springer-Verlag, 1984.
P. Jordan, J. von Neumann and E. Wigner, On an algebraic generalization of the quantum mechanical formalism, Annals of Mathematics 35 (1934). Formal reality outside associative algebra.
J. Bochnak, M. Coste and M.-F. Roy, Real Algebraic Geometry, Ergebnisse der Mathematik 36, Springer-Verlag, 1998.

AI Suggested Questions

For which classes of noncommutative domains is failure of formal reality equivalent to $- 1$ lying in the weak preordering?
How does R. E. Johnson's theorem interact with Ore localisation and with rings that do not embed into division rings?
What is the noncommutative analogue of the real spectrum, and does it carry a useful topology?
Which finitely presented algebras can be shown formally real by an explicit ordering rather than by an abstract argument?
How does formal reality for Jordan algebras relate to formal reality for their associative enveloping algebras?
What degree bounds are known for representing a totally positive polynomial as an element of the weak preordering?
Can formal reality be characterised by the absence of a specific finite family of identities, uniformly in the ring?

01Executive Summary

02Overview

03Learning Objectives

04Definitions

05Core Concepts

Why T(R) is the right generating set

The commutative shadow

What formal reality already implies

06Key Results

07Proof Techniques and Method

Generate, then test properness

Maximise by Zorn, then use maximality

Invert inside the cone

08Worked Example

Formally real, and not

A noncommutative formally real ring

A calculation in T(R)

Totally positive is not the same as in T(R)

09Comparison and Classification

10Relationship Map

11Applications and Industry Use

Real closed fields and model theory

Formally real Jordan algebras

Sums-of-squares relaxations

Zero divisors and units

12Standards and Notation

13Failure Modes and Common Mistakes

14Historical Notes and Lessons Learned

15Quick Reference

16Frequently Asked Questions

17Related KEVOS Topics

18References

19AI Suggested Questions

Continue learning

Executive Summary

Overview

Learning Objectives

Definitions

Core Concepts

Why $T (R)$ is the right generating set

Key Results

Proof Techniques and Method

Worked Example

A calculation in $T (R)$

Totally positive is not the same as in $T (R)$

Comparison and Classification

Relationship Map

Applications and Industry Use

Standards and Notation

Failure Modes and Common Mistakes

Historical Notes and Lessons Learned

Quick Reference

Frequently Asked Questions

Related KEVOS Topics

References

AI Suggested Questions