The lattice of linear Malcev conditions Jakub Opr sal Charles - - PowerPoint PPT Presentation

the lattice of linear mal cev conditions
SMART_READER_LITE
LIVE PREVIEW

The lattice of linear Malcev conditions Jakub Opr sal Charles - - PowerPoint PPT Presentation

The lattice of linear Malcev conditions Jakub Opr sal Charles University in Prague Nashville, May 29, 2015 Posets of Malcev conditions and interpretability types A Malcev condition is a condition of the form there exists some terms


slide-1
SLIDE 1

The lattice of linear Mal’cev conditions

Jakub Oprˇ sal

Charles University in Prague

Nashville, May 29, 2015

slide-2
SLIDE 2

Posets of Mal’cev conditions and interpretability types

A Mal’cev condition is a condition of the form there exists some terms satisfying some equations. Mal’cev conditions are naturally ordered by implication. A stronger condition is larger then a weaker one.

slide-3
SLIDE 3

Posets of Mal’cev conditions and interpretability types

A Mal’cev condition is a condition of the form there exists some terms satisfying some equations. Mal’cev conditions are naturally ordered by implication. A stronger condition is larger then a weaker one. A clone homomorphism (or interpretation) from a clone A to a clone B is a map i : A → B mapping n-ary operations to n-operations, and preserving composition and projections. Interpretation from a variety V to a variety W is a functor I : W → V that is commuting with forgetful functors.

slide-4
SLIDE 4

Posets of Mal’cev conditions and interpretability types

A Mal’cev condition is a condition of the form there exists some terms satisfying some equations. Mal’cev conditions are naturally ordered by implication. A stronger condition is larger then a weaker one. A clone homomorphism (or interpretation) from a clone A to a clone B is a map i : A → B mapping n-ary operations to n-operations, and preserving composition and projections. Interpretation from a variety V to a variety W is a functor I : W → V that is commuting with forgetful functors. Interpretability form quasi-order. By a standard technique, we can get the corresponding partial order (we factor by equi-interpretability). (Garcia, Taylor: The lattice of interpretability types of varieties, 1984.)

slide-5
SLIDE 5

Joins

Join of two Mal’cev conditions is the condition given by conjuction of the two.

slide-6
SLIDE 6

Joins

Join of two Mal’cev conditions is the condition given by conjuction of the two. Join of two varieties V and W in can be described as the variety V ∨ W whose operations are operations of both varieties (taken as a discrete union of operations of V and operations W), and whose identities are all identities of both varieties.

slide-7
SLIDE 7

Joins

Join of two Mal’cev conditions is the condition given by conjuction of the two. Join of two varieties V and W in can be described as the variety V ∨ W whose operations are operations of both varieties (taken as a discrete union of operations of V and operations W), and whose identities are all identities of both varieties. In the other worlds, we can describe algebras in V ∨ W as (A, F ∪ G) where (A, F) ∈ V and (A, G) ∈ W.

slide-8
SLIDE 8

Joins

Join of two Mal’cev conditions is the condition given by conjuction of the two. Join of two varieties V and W in can be described as the variety V ∨ W whose operations are operations of both varieties (taken as a discrete union of operations of V and operations W), and whose identities are all identities of both varieties. In the other worlds, we can describe algebras in V ∨ W as (A, F ∪ G) where (A, F) ∈ V and (A, G) ∈ W.

Examples

◮ Mal’cev ∨ J´

  • nsson terms = Pixley term,

◮ J´

  • nsson terms ∨ cube term = near unanimity.

◮ Gumm terms ∨ SD(∨) = J´

  • nsson terms.
slide-9
SLIDE 9

Meets

Meet of two abstract clones A and B is a clone A × B (the product in the category of clones) that is described by (A × B)[n] = A[n] × B[n] with the obvious composition, and obvious projections.

slide-10
SLIDE 10

Meets

Meet of two abstract clones A and B is a clone A × B (the product in the category of clones) that is described by (A × B)[n] = A[n] × B[n] with the obvious composition, and obvious projections. For varieties V1 and V2 the meet is described as the variety V1 × V2 that is defined in such a way that

  • 1. its signature is disjoint union of signtures of V1 and W with a new

binary symbol ·,

slide-11
SLIDE 11

Meets

Meet of two abstract clones A and B is a clone A × B (the product in the category of clones) that is described by (A × B)[n] = A[n] × B[n] with the obvious composition, and obvious projections. For varieties V1 and V2 the meet is described as the variety V1 × V2 that is defined in such a way that

  • 1. its signature is disjoint union of signtures of V1 and W with a new

binary symbol ·,

  • 2. it has two subvarieties V′

1 and V′ 2 that are equi-interpretable with V1,

V2 respectively (Vi satisfies x1 · x2 ≈ xi),

slide-12
SLIDE 12

Meets

Meet of two abstract clones A and B is a clone A × B (the product in the category of clones) that is described by (A × B)[n] = A[n] × B[n] with the obvious composition, and obvious projections. For varieties V1 and V2 the meet is described as the variety V1 × V2 that is defined in such a way that

  • 1. its signature is disjoint union of signtures of V1 and W with a new

binary symbol ·,

  • 2. it has two subvarieties V′

1 and V′ 2 that are equi-interpretable with V1,

V2 respectively (Vi satisfies x1 · x2 ≈ xi),

  • 3. every algebra in V1 × V2 is a product of an algebra from V′

1 and

an algebra from V′

2.

slide-13
SLIDE 13

Poset of linear Mal’cev conditions

A linear Mal’cev condition is a condition that do not include term composition, i.e., only equations of the form f (xi1, . . . , xin) ≈ g(xj1, . . . , xim),

  • r

f (xi1, . . . , xin) ≈ xj are allowed.

slide-14
SLIDE 14

Poset of linear Mal’cev conditions

A linear Mal’cev condition is a condition that do not include term composition, i.e., only equations of the form f (xi1, . . . , xin) ≈ g(xj1, . . . , xim),

  • r

f (xi1, . . . , xin) ≈ xj are allowed.

Examples

Mal’cev term, Pixley term, Day terms, Gumm terms, near unanimity, cube term, J´

  • nsson terms, etc.
slide-15
SLIDE 15

Poset of linear Mal’cev conditions

A linear Mal’cev condition is a condition that do not include term composition, i.e., only equations of the form f (xi1, . . . , xin) ≈ g(xj1, . . . , xim),

  • r

f (xi1, . . . , xin) ≈ xj are allowed.

Examples

Mal’cev term, Pixley term, Day terms, Gumm terms, near unanimity, cube term, J´

  • nsson terms, etc.

Not examples

group terms, lattice terms, semilattice term, congruence uniformity, congruence singularity?.

slide-16
SLIDE 16

Poset of linear Mal’cev conditions

A linear Mal’cev condition is a condition that do not include term composition, i.e., only equations of the form f (xi1, . . . , xin) ≈ g(xj1, . . . , xim),

  • r

f (xi1, . . . , xin) ≈ xj are allowed.

Examples

Mal’cev term, Pixley term, Day terms, Gumm terms, near unanimity, cube term, J´

  • nsson terms, etc.

Not examples

group terms, lattice terms, semilattice term, congruence uniformity, congruence singularity?. Linear Mal’cev condition forms a subposet of the lattice of all Mal’cev conditions.

slide-17
SLIDE 17

Poset of linear Mal’cev conditions

A linear Mal’cev condition is a condition that do not include term composition, i.e., only equations of the form f (xi1, . . . , xin) ≈ g(xj1, . . . , xim),

  • r

f (xi1, . . . , xin) ≈ xj are allowed.

Examples

Mal’cev term, Pixley term, Day terms, Gumm terms, near unanimity, cube term, J´

  • nsson terms, etc.

Not examples

group terms, lattice terms, semilattice term, congruence uniformity, congruence singularity?. Linear Mal’cev condition forms a subposet of the lattice of all Mal’cev conditions. But, the subposet is not a sublattice!

slide-18
SLIDE 18

Meet of linear conditions

Proposition

Meet of Mal’cev term and congruence distributivity is not equivalent to any linear Mal’cev condition.

slide-19
SLIDE 19

Meet of linear conditions

Proposition

Meet of Mal’cev term and congruence distributivity is not equivalent to any linear Mal’cev condition.

Definition (Barto, Pinsker)

An algebra A is said to be a retract of B if there are two maps a: B → A and b: A → B such that ab = 1A, and for every basic operation f we have fA(a1, . . . , an) = afB(b(a1), . . . , b(an)).

slide-20
SLIDE 20

Meet of linear conditions

Proposition

Meet of Mal’cev term and congruence distributivity is not equivalent to any linear Mal’cev condition.

Definition (Barto, Pinsker)

An algebra A is said to be a retract of B if there are two maps a: B → A and b: A → B such that ab = 1A, and for every basic operation f we have fA(a1, . . . , an) = afB(b(a1), . . . , b(an)).

Observation

If A is a retract of B then A satisfies all the linear equations that B does.

slide-21
SLIDE 21

Meet of linear conditions (cont.)

We will show that meet of Mal’cev and majority is not linear.

slide-22
SLIDE 22

Meet of linear conditions (cont.)

We will show that meet of Mal’cev and majority is not linear. Let

◮ V1 be the variety with single ternary Mal’cev operation q,

slide-23
SLIDE 23

Meet of linear conditions (cont.)

We will show that meet of Mal’cev and majority is not linear. Let

◮ V1 be the variety with single ternary Mal’cev operation q, ◮ V2 be the variety with the majority operation m,

slide-24
SLIDE 24

Meet of linear conditions (cont.)

We will show that meet of Mal’cev and majority is not linear. Let

◮ V1 be the variety with single ternary Mal’cev operation q, ◮ V2 be the variety with the majority operation m, ◮ W a variety equi-interpretable with V1 × V2 that is defined by linear

equations.

slide-25
SLIDE 25

Meet of linear conditions (cont.)

We will show that meet of Mal’cev and majority is not linear. Let

◮ V1 be the variety with single ternary Mal’cev operation q, ◮ V2 be the variety with the majority operation m, ◮ W a variety equi-interpretable with V1 × V2 that is defined by linear

equations. We choose algebra in V′

1 that has no J´

  • nsson terms, and similarly algebra

in V′

2 that has no Mal’cev term.

slide-26
SLIDE 26

Meet of linear conditions (cont.)

We will show that meet of Mal’cev and majority is not linear. Let

◮ V1 be the variety with single ternary Mal’cev operation q, ◮ V2 be the variety with the majority operation m, ◮ W a variety equi-interpretable with V1 × V2 that is defined by linear

equations. We choose algebra in V′

1 that has no J´

  • nsson terms, and similarly algebra

in V′

2 that has no Mal’cev term. For example ◮ A = ({0, 1}, x + y + z, proj3 1, proj2 1), and ◮ B = ({0, 1}, proj3 1, (x ∨ y) ∧ (y ∨ z) ∧ (x ∨ z), proj2 2).

slide-27
SLIDE 27

Meet of linear conditions (cont.)

Consider the interpretation of A × B in W, and take its retract C via (0,0) (0,1) (1,0) (1,1) 1 2

slide-28
SLIDE 28

Meet of linear conditions (cont.)

Consider the interpretation of A × B in W, and take its retract C via (0,0) (0,1) (1,0) (1,1) 1 2 Finally, let C′ be the interpretation of C in V1 × V2.

slide-29
SLIDE 29

Meet of linear conditions (cont.)

Consider the interpretation of A × B in W, and take its retract C via (0,0) (0,1) (1,0) (1,1) 1 2 Finally, let C′ be the interpretation of C in V1 × V2. Then

  • 1. Both B′ = {0, 1} and A′ = {1, 2} are subuniverses of C′, Clo A′ is

a reduct of Clo A, and Clo B′ is a reduct of Clo B,

slide-30
SLIDE 30

Meet of linear conditions (cont.)

Consider the interpretation of A × B in W, and take its retract C via (0,0) (0,1) (1,0) (1,1) 1 2 Finally, let C′ be the interpretation of C in V1 × V2. Then

  • 1. Both B′ = {0, 1} and A′ = {1, 2} are subuniverses of C′, Clo A′ is

a reduct of Clo A, and Clo B′ is a reduct of Clo B,

  • 2. |C ′| = 3 which is a prime! So, either C′ ∈ V′

1, or C′ ∈ V′ 2,

slide-31
SLIDE 31

Meet of linear conditions (cont.)

Consider the interpretation of A × B in W, and take its retract C via (0,0) (0,1) (1,0) (1,1) 1 2 Finally, let C′ be the interpretation of C in V1 × V2. Then

  • 1. Both B′ = {0, 1} and A′ = {1, 2} are subuniverses of C′, Clo A′ is

a reduct of Clo A, and Clo B′ is a reduct of Clo B,

  • 2. |C ′| = 3 which is a prime! So, either C′ ∈ V′

1, or C′ ∈ V′ 2,

  • 3. but neither is possible since A has no majority term, and B has no

Mal’cev term!

slide-32
SLIDE 32

Meet of linear conditions (cont.)

Consider the interpretation of A × B in W, and take its retract C via (0,0) (0,1) (1,0) (1,1) 1 2 Finally, let C′ be the interpretation of C in V1 × V2. Then

  • 1. Both B′ = {0, 1} and A′ = {1, 2} are subuniverses of C′, Clo A′ is

a reduct of Clo A, and Clo B′ is a reduct of Clo B,

  • 2. |C ′| = 3 which is a prime! So, either C′ ∈ V′

1, or C′ ∈ V′ 2,

  • 3. but neither is possible since A has no majority term, and B has no

Mal’cev term!

slide-33
SLIDE 33

Lattice of linear varieties

Problems with Mal’cev conditions

slide-34
SLIDE 34

Lattice of linear varieties

Problems with Mal’cev conditions

◮ they are not closed under infinite joins,

slide-35
SLIDE 35

Lattice of linear varieties

Problems with Mal’cev conditions

◮ they are not closed under infinite joins, ◮ there is not a largest linear Mal’cev condition interpretable in some

clone (or non-linear Mal’cev condition).

slide-36
SLIDE 36

Lattice of linear varieties

Problems with Mal’cev conditions

◮ they are not closed under infinite joins, ◮ there is not a largest linear Mal’cev condition interpretable in some

clone (or non-linear Mal’cev condition). These problems can be solved by taking all linear varieties instead. (We lose Mal’cev conditions that are not strong.)

slide-37
SLIDE 37

Prime elements of the lattice

slide-38
SLIDE 38

Prime elements of the lattice

(Sequeira, Barto) Let X be a given set of variables, and A ⊆ Eq(X). We say that variety V is A-colorable if there is a map c : FV(X) → X such that

  • 1. c(x) = x for all x ∈ X, and
  • 2. for every α ∈ A whenever f ∼ˆ

α g then c(f ) ∼α c(g)

where ˆ α denotes the congruence of the free algebra over X generated by α.

slide-39
SLIDE 39

Prime elements of the lattice

(Sequeira, Barto) Let X be a given set of variables, and A ⊆ Eq(X). We say that variety V is A-colorable if there is a map c : FV(X) → X such that

  • 1. c(x) = x for all x ∈ X, and
  • 2. for every α ∈ A whenever f ∼ˆ

α g then c(f ) ∼α c(g)

where ˆ α denotes the congruence of the free algebra over X generated by α. We say that Mal’cev condition P satisfies coloring condition A if variety V satisfies P if and only if V is not A-colorable.

slide-40
SLIDE 40

Prime elements of the lattice

(Sequeira, Barto) Let X be a given set of variables, and A ⊆ Eq(X). We say that variety V is A-colorable if there is a map c : FV(X) → X such that

  • 1. c(x) = x for all x ∈ X, and
  • 2. for every α ∈ A whenever f ∼ˆ

α g then c(f ) ∼α c(g)

where ˆ α denotes the congruence of the free algebra over X generated by α. We say that Mal’cev condition P satisfies coloring condition A if variety V satisfies P if and only if V is not A-colorable. Many of Mal’cev conditions that are suspected to be prime satisfy some coloring condition. Namely

slide-41
SLIDE 41

Prime elements of the lattice

(Sequeira, Barto) Let X be a given set of variables, and A ⊆ Eq(X). We say that variety V is A-colorable if there is a map c : FV(X) → X such that

  • 1. c(x) = x for all x ∈ X, and
  • 2. for every α ∈ A whenever f ∼ˆ

α g then c(f ) ∼α c(g)

where ˆ α denotes the congruence of the free algebra over X generated by α. We say that Mal’cev condition P satisfies coloring condition A if variety V satisfies P if and only if V is not A-colorable. Many of Mal’cev conditions that are suspected to be prime satisfy some coloring condition. Namely

◮ congruence n-permutability, ◮ congruence modularity, ◮ satisfying non-trivial congruence identity, ◮ n-cube terms, ◮ triviality (x ≈ y).

slide-42
SLIDE 42

Prime elements of the lattice (cont.)

Theorem (Sequeira; Bentz-Sequeira)

Congruence modularity, n-permutability, satisfying non-trivial congruence identity, and n-cube term are prime with respect to varieties axiomatized by linear equations.

slide-43
SLIDE 43

Prime elements of the lattice (cont.)

Theorem (Sequeira; Bentz-Sequeira)

Congruence modularity, n-permutability, satisfying non-trivial congruence identity, and n-cube term are prime with respect to varieties axiomatized by linear equations.

Theorem (O.)

A Mal’cev condition M satisfy coloring condition A if and only if for every linear variety V we have that either V satisfies M, or V is interpretable in Pol(X, A)

slide-44
SLIDE 44

Prime elements of the lattice (cont.)

Theorem (Sequeira; Bentz-Sequeira)

Congruence modularity, n-permutability, satisfying non-trivial congruence identity, and n-cube term are prime with respect to varieties axiomatized by linear equations.

Theorem (O.)

A Mal’cev condition M satisfy coloring condition A if and only if for every linear variety V we have that either V satisfies M, or V is interpretable in Pol(X, A)

Proof.

slide-45
SLIDE 45

Prime elements of the lattice (cont.)

Theorem (Sequeira; Bentz-Sequeira)

Congruence modularity, n-permutability, satisfying non-trivial congruence identity, and n-cube term are prime with respect to varieties axiomatized by linear equations.

Theorem (O.)

A Mal’cev condition M satisfy coloring condition A if and only if for every linear variety V we have that either V satisfies M, or V is interpretable in Pol(X, A)

Proof.

Suppose that V is linear and A-colorable (A ⊆ Eq X). Then we define an interpretation i : V → Pol(X, A) as i(f )(x0, . . . , xn) = c(f (x0, x1, . . . , xn)) for every basic operation f , and extend to terms.

slide-46
SLIDE 46

Prime elements of the lattice (cont.)

Theorem (Sequeira; Bentz-Sequeira)

Congruence modularity, n-permutability, satisfying non-trivial congruence identity, and n-cube term are prime with respect to varieties axiomatized by linear equations.

Theorem (O.)

A Mal’cev condition M satisfy coloring condition A if and only if for every linear variety V we have that either V satisfies M, or V is interpretable in Pol(X, A)

Proof.

Suppose that V is linear and A-colorable (A ⊆ Eq X). Then we define an interpretation i : V → Pol(X, A) as i(f )(x0, . . . , xn) = c(f (x0, x1, . . . , xn)) for every basic operation f , and extend to terms.

slide-47
SLIDE 47

Splitting of other lattices

We say that two subsets of elements A, and B split a lattice if for every element x of the lattice we have either a ≤ x for some a ∈ A, or x ≤ b for some b ∈ B.

slide-48
SLIDE 48

Splitting of other lattices

We say that two subsets of elements A, and B split a lattice if for every element x of the lattice we have either a ≤ x for some a ∈ A, or x ≤ b for some b ∈ B.

Theorem (Valeriote, Willard, 2014)

n-permutability and idempotent polymorhisms of two-element poset split the lattice of idempotent varieties.

slide-49
SLIDE 49

Splitting of other lattices

We say that two subsets of elements A, and B split a lattice if for every element x of the lattice we have either a ≤ x for some a ∈ A, or x ≤ b for some b ∈ B.

Theorem (Valeriote, Willard, 2014)

n-permutability and idempotent polymorhisms of two-element poset split the lattice of idempotent varieties.

Theorem (Kiss, Kearnes, 2013)

Satisfying a non-trivial congruence identity and the set {Pol(L) : L is a semilattice} split the lattice of idempotent varieties.

slide-50
SLIDE 50

Some open problems. . .

Problem

Find a satisfactory description of linear meet.

slide-51
SLIDE 51

Some open problems. . .

Problem

Find a satisfactory description of linear meet.

Problem

Is CM the linear meet of Mal’cev and CD?

slide-52
SLIDE 52

Some open problems. . .

Problem

Find a satisfactory description of linear meet.

Problem

Is CM the linear meet of Mal’cev and CD?

Problem

Does every prime element of the linear lattice satisfy some coloring condition? (Need to consider a little generalized conditions.)

slide-53
SLIDE 53

Some open problems. . .

Problem

Find a satisfactory description of linear meet.

Problem

Is CM the linear meet of Mal’cev and CD?

Problem

Does every prime element of the linear lattice satisfy some coloring condition? (Need to consider a little generalized conditions.)

Problem

Are the Mal’cev conditions that satisfy some coloring condition prime? (Known for Mal’cev, cyclic terms, not known for everything else.)

slide-54
SLIDE 54

Some open problems. . .

Problem

Find a satisfactory description of linear meet.

Problem

Is CM the linear meet of Mal’cev and CD?

Problem

Does every prime element of the linear lattice satisfy some coloring condition? (Need to consider a little generalized conditions.)

Problem

Are the Mal’cev conditions that satisfy some coloring condition prime? (Known for Mal’cev, cyclic terms, not known for everything else.) Thank you for your attention!