The Possible Structure of the Mitchell Order Omer Ben-Neria UCLA - - PowerPoint PPT Presentation

the possible structure of the mitchell order
SMART_READER_LITE
LIVE PREVIEW

The Possible Structure of the Mitchell Order Omer Ben-Neria UCLA - - PowerPoint PPT Presentation

The Possible Structure of the Mitchell Order Omer Ben-Neria UCLA HIFW02, University of East Anglia, November 2015 Definitions 1. In this talk: Order = Partial ordered set. 2. A normal measure U on is a complete normal ultrafilter on


slide-1
SLIDE 1

The Possible Structure of the Mitchell Order

Omer Ben-Neria

UCLA

HIFW02, University of East Anglia, November 2015

slide-2
SLIDE 2

Definitions

  • 1. In this talk: Order = Partial ordered set.
  • 2. A normal measure U on κ is a κ−complete normal

ultrafilter on κ.

  • 3. U ⊳ W ⇐

⇒ U ∈ Ult(V , W )

  • 4. ⊳(κ) is the restriction of ⊳ to the set of normal

measures on κ.

  • 5. o(κ) = rank(⊳(κ)) (⊳(κ) is well-founded)
  • 6. An order (S, <S) is realized as ⊳(κ) in M if

(S, <S) ∼ = ⊳(κ)M.

slide-3
SLIDE 3

Definitions

  • 1. In this talk: Order = Partial ordered set.
  • 2. A normal measure U on κ is a κ−complete normal

ultrafilter on κ.

  • 3. U ⊳ W ⇐

⇒ U ∈ Ult(V , W )

  • 4. ⊳(κ) is the restriction of ⊳ to the set of normal

measures on κ.

  • 5. o(κ) = rank(⊳(κ)) (⊳(κ) is well-founded)
  • 6. An order (S, <S) is realized as ⊳(κ) in M if

(S, <S) ∼ = ⊳(κ)M. Goal : Determine what are the well-founded orders that can be realized as ⊳(κ)

slide-4
SLIDE 4

The Possible Number of Normal Measures on κ

The number of normal measures on κ = | ⊳ (κ)|. Author Possible | ⊳ (κ)| Assumption Kunen 1 minimal Kunen-Paris κ++ minimal Mitchell any λ ≤ κ++

  • (κ) = λ

Baldwin any λ < κ 1

  • (κ) >> λ

Apter-Cummings-Hamkins κ+ minimal Leaning any λ < κ+ less than

  • (κ) = 2

Friedman-Magidor any λ ≤ κ++ minimal

1κ is also the first measurable cardinal

slide-5
SLIDE 5

Previous Results on the possible structure of ⊳(κ): Authors Possible to realize as ⊳(κ) Mitchell well-orders Baldwin pre-well-orders Cummings Large orders, embed every tame order up to a certain rank Witzany Large orders, embed every well-founded order of size ≤ κ+

slide-6
SLIDE 6

“... it is not known whether o(κ) = ω implies that there is a coherent sequence U of measures in V with oU(κ) = ω.”

(William J. Mitchell - Handbook of Set Theory/Beginning Inner Model Theory)

slide-7
SLIDE 7

For a negative answer, we want to realize the following order

1 2

  • (0,0)

B0

(1,1)

  • (1,2)
  • B1

(2,3)

  • (2,4)
  • (2,5)
  • B2

...... n ............

  • (n,kn)
  • (n,kn)

. . . . . . . . . . . .

  • (n,kn+n)

Bn

slide-8
SLIDE 8

Results

Part I The Orders - Tame orders The Result - Tame orders of cardinality ≤ κ can be realized as ⊳(κ) from assumptions weaker than o(κ) = κ+.

slide-9
SLIDE 9

Results

Part I The Orders - Tame orders The Result - Tame orders of cardinality ≤ κ can be realized as ⊳(κ) from assumptions weaker than o(κ) = κ+. Part II The Orders - Arbitrary well-founded orders The Result - Well-founded orders of cardinality ≤ κ can be realized as ⊳(κ) from assumptions slightly stronger than 0¶

slide-10
SLIDE 10

Part I

Tame orders

slide-11
SLIDE 11

Part I - Tame Orders (1/3)

A well-founded order is Tame if it does not embed two specific

  • rders R2,2 and Sω,2.

R2,2 = {x0, y0, x1, y1}, <R2,2= {(x0, y0), (x1, y1)}

  • x0
  • y0
  • x1
  • y1
slide-12
SLIDE 12

Part I - Tame Orders (1/3)

A well-founded order is Tame if it does not embed two specific

  • rders R2,2 and Sω,2.

R2,2 = {x0, y0, x1, y1}, <R2,2= {(x0, y0), (x1, y1)}

  • x0
  • y0
  • x1
  • y1

Sω,2 = {xn}n<ω ⊎ {yn}n<ω, <Sω,2= {(xn′, yn) | n′ ≥ n}

  • x0
  • x1
  • x2

. . . . . . . . .

  • xn

. . . . . .

  • y0
  • y1
  • y2 . . . . . . . . .
  • yn . . . . . .
slide-13
SLIDE 13

Part I - Tame Orders (2/3)

Suppose (S, <S) is an order. For every x ∈ S let u(x) = {y ∈ S | x <S y}, and define U(S) = {u(x) | x ∈ S}

slide-14
SLIDE 14

Part I - Tame Orders (2/3)

Suppose (S, <S) is an order. For every x ∈ S let u(x) = {y ∈ S | x <S y}, and define U(S) = {u(x) | x ∈ S}

◮ If (S, <S) does not embed R2,2 then for every x, x′ ∈ S, u(x),

u(x′) are ⊆ −comparable.

slide-15
SLIDE 15

Part I - Tame Orders (2/3)

Suppose (S, <S) is an order. For every x ∈ S let u(x) = {y ∈ S | x <S y}, and define U(S) = {u(x) | x ∈ S}

◮ If (S, <S) does not embed R2,2 then for every x, x′ ∈ S, u(x),

u(x′) are ⊆ −comparable. Otherwise: <S↾ {x, y, x′, y′} ≃ R2,2 for some y, y′.

  • x
  • x′

u(x) u(x′)

  • y
  • y′
slide-16
SLIDE 16

Part I - Tame Orders (3/3)

◮ If (S, <S) does not embed R2,2 then (U(S), ⊃) is a linear

  • rdering.
slide-17
SLIDE 17

Part I - Tame Orders (3/3)

◮ If (S, <S) does not embed R2,2 then (U(S), ⊃) is a linear

  • rdering.

◮ If (S, <S) does not embed Sω,2 as well then (U(S), ⊃) is a

well-order.

slide-18
SLIDE 18

Part I - Tame Orders (3/3)

◮ If (S, <S) does not embed R2,2 then (U(S), ⊃) is a linear

  • rdering.

◮ If (S, <S) does not embed Sω,2 as well then (U(S), ⊃) is a

well-order.

◮ For every tame order (S, <S) we define the tame rank of

(S, <S): Trank(S, <S) = otp(U(S), ⊃)

slide-19
SLIDE 19

Part I - Tame Orders (3/3)

◮ If (S, <S) does not embed R2,2 then (U(S), ⊃) is a linear

  • rdering.

◮ If (S, <S) does not embed Sω,2 as well then (U(S), ⊃) is a

well-order.

◮ For every tame order (S, <S) we define the tame rank of

(S, <S): Trank(S, <S) = otp(U(S), ⊃)

rank(S, <S) ≤ Trank(S, <S) < |S|+

slide-20
SLIDE 20

Part I - Main Result

Theorem 1 (BN) Suppose κ is measurable in V and (S, <S) ∈ V is a tame

  • rder such that

◮ |S| ≤ κ and ◮ Trank(S, <S) ≤ oV (κ),

then (S, <S) can be realized as ⊳(κ) in a cofinality preserving extension.

slide-21
SLIDE 21

Part I - Example

◮ Let S2,2 = {x0, y0, x1, y1}, <S2,2= {(x0, y0), (x1, y1), (x1, y0)}

  • x0
  • y0
  • x1
  • y1
slide-22
SLIDE 22

Part I - Example

◮ Let S2,2 = {x0, y0, x1, y1}, <S2,2= {(x0, y0), (x1, y1), (x1, y0)}

  • x0
  • y0
  • x1
  • y1

◮ Trank(S2,2) = 3,

z y0, y1 x0 x1 u(z) ∅ {y0} {y0, y1}

slide-23
SLIDE 23

Part I - Example

◮ Let S2,2 = {x0, y0, x1, y1}, <S2,2= {(x0, y0), (x1, y1), (x1, y0)}

  • x0
  • y0
  • x1
  • y1

◮ Trank(S2,2) = 3,

z y0, y1 x0 x1 u(z) ∅ {y0} {y0, y1}

◮ Can realize S2,2 as ⊳(κ) from o(κ) = 3

slide-24
SLIDE 24

Principal non-tame orders

slide-25
SLIDE 25

Principal non-tame orders

slide-26
SLIDE 26

Part II

Goal: Realizing arbitrary well-founded orders starting from models with overlapping extenders

slide-27
SLIDE 27

First realize R2,2 and Sω,2 (3 steps):

slide-28
SLIDE 28

First realize R2,2 and Sω,2 (3 steps):

  • 1. Describe the ground model assumptions V = L[E] and

Introduce the extenders Fα,n

slide-29
SLIDE 29

First realize R2,2 and Sω,2 (3 steps):

  • 1. Describe the ground model assumptions V = L[E] and

Introduce the extenders Fα,n

  • 2. Force over V with an iteration of a Collapsing and Coding

posets, replace Fα,n with Uα,n

slide-30
SLIDE 30

First realize R2,2 and Sω,2 (3 steps):

  • 1. Describe the ground model assumptions V = L[E] and

Introduce the extenders Fα,n

  • 2. Force over V with an iteration of a Collapsing and Coding

posets, replace Fα,n with Uα,n

  • 3. Use Uα,n to realize Sω,2 and R2,2
slide-31
SLIDE 31

Part II - Ground Model Assumptions

Suppose that V = L[E] be an extender model where

  • 1. κ < θ are measurable, θ is the first measurable above κ
slide-32
SLIDE 32

Part II - Ground Model Assumptions

Suppose that V = L[E] be an extender model where

  • 1. κ < θ are measurable, θ is the first measurable above κ
  • 2. There is a ⊳−increasing sequence

F = Fα | α < λ of (κ, θ++)−extenders, λ < θ.

slide-33
SLIDE 33

Part II - Ground Model Assumptions

Suppose that V = L[E] be an extender model where

  • 1. κ < θ are measurable, θ is the first measurable above κ
  • 2. There is a ⊳−increasing sequence

F = Fα | α < λ of (κ, θ++)−extenders, λ < θ.

  • 3. Vθ+2 ⊂ Ult(V , Fα) for every α < λ
slide-34
SLIDE 34

Part II - Ground Model Assumptions

Suppose that V = L[E] be an extender model where

  • 1. κ < θ are measurable, θ is the first measurable above κ
  • 2. There is a ⊳−increasing sequence

F = Fα | α < λ of (κ, θ++)−extenders, λ < θ.

  • 3. Vθ+2 ⊂ Ult(V , Fα) for every α < λ

4. F consists of all the full (κ, θ++)−extenders on E

  • 5. There are no stronger extenders on κ in E
slide-35
SLIDE 35

Part II - Ground Model Assumptions

Suppose that V = L[E] be an extender model where

  • 1. κ < θ are measurable, θ is the first measurable above κ
  • 2. There is a ⊳−increasing sequence

F = Fα | α < λ of (κ, θ++)−extenders, λ < θ.

  • 3. Vθ+2 ⊂ Ult(V , Fα) for every α < λ

4. F consists of all the full (κ, θ++)−extenders on E

  • 5. There are no stronger extenders on κ in E

θ has a unique normal measure Uθ in V , Uθ ∈ Vθ+2, so Uθ ⊳ Fα for every α < λ

slide-36
SLIDE 36

Part II - The extenders Fα,n (1/3)

For every n < ω define

◮ in : V → Mn = Ult(n)(V , Uθ) the n−th iterated ultrapower of

V by Uθ.

◮ θn = in(θ) > θ, is the first measurable cardinal above κ in Mn.

slide-37
SLIDE 37

Part II - The extenders Fα,n (1/3)

For every n < ω define

◮ in : V → Mn = Ult(n)(V , Uθ) the n−th iterated ultrapower of

V by Uθ.

◮ θn = in(θ) > θ, is the first measurable cardinal above κ in Mn. ◮ Note that θ++ is a fixed point of in and θ++ = (θ++ n

)Mn.

slide-38
SLIDE 38

Part II - The extenders Fα,n (1/3)

For every n < ω define

◮ in : V → Mn = Ult(n)(V , Uθ) the n−th iterated ultrapower of

V by Uθ.

◮ θn = in(θ) > θ, is the first measurable cardinal above κ in Mn. ◮ Note that θ++ is a fixed point of in and θ++ = (θ++ n

)Mn.

◮ Fα,n = in(Fα) is a (κ, θ++V )−extender for Mn and V .

slide-39
SLIDE 39

Part II - The extenders Fα,n (1/3)

For every n < ω define

◮ in : V → Mn = Ult(n)(V , Uθ) the n−th iterated ultrapower of

V by Uθ.

◮ θn = in(θ) > θ, is the first measurable cardinal above κ in Mn. ◮ Note that θ++ is a fixed point of in and θ++ = (θ++ n

)Mn.

◮ Fα,n = in(Fα) is a (κ, θ++V )−extender for Mn and V . ◮ θn is the first measurable cardinal above κ in Ult(V , Fα,n)

slide-40
SLIDE 40

Part II - The extenders Fα,n (2/3)

Suppose α′ < α < λ then Fα′ ⊳ Fα so

slide-41
SLIDE 41

Part II - The extenders Fα,n (2/3)

Suppose α′ < α < λ then Fα′ ⊳ Fα so

◮ Fα′,1 ⊳ Fα,1,

slide-42
SLIDE 42

Part II - The extenders Fα,n (2/3)

Suppose α′ < α < λ then Fα′ ⊳ Fα so

◮ Fα′,1 ⊳ Fα,1, ◮ If n > 1 then Fα′,n = i1,n(Fα′,1) ⊳ Fα,1

slide-43
SLIDE 43

Part II - The extenders Fα,n (2/3)

Suppose α′ < α < λ then Fα′ ⊳ Fα so

◮ Fα′,1 ⊳ Fα,1, ◮ If n > 1 then Fα′,n = i1,n(Fα′,1) ⊳ Fα,1 ◮ Fα′,0 ⊳ Fα,1

slide-44
SLIDE 44

Part II - The extenders Fα,n (2/3)

Suppose α′ < α < λ then Fα′ ⊳ Fα so

◮ Fα′,1 ⊳ Fα,1, ◮ If n > 1 then Fα′,n = i1,n(Fα′,1) ⊳ Fα,1 ◮ Fα′,0 ⊳ Fα,1

  • 1. Uθ ∈ Ult(Vκ+1, Fα′,0)
  • 2. if Fα′,0 ⊳ Fα,1 then Uθ ∈ Ult(V , Fα,1)
  • 3. impossible as θ1 > θ is the first measurable cardinal above κ in

Ult(V , Fα,1)

slide-45
SLIDE 45

Part II - The extenders Fα,n (2/3)

Suppose α′ < α < λ then Fα′ ⊳ Fα so

◮ Fα′,1 ⊳ Fα,1, ◮ If n > 1 then Fα′,n = i1,n(Fα′,1) ⊳ Fα,1 ◮ Fα′,0 ⊳ Fα,1

  • 1. Uθ ∈ Ult(Vκ+1, Fα′,0)
  • 2. if Fα′,0 ⊳ Fα,1 then Uθ ∈ Ult(V , Fα,1)
  • 3. impossible as θ1 > θ is the first measurable cardinal above κ in

Ult(V , Fα,1)

◮ Conclusion: Fα′,n′ ⊳ Fα,1 iff n′ ≥ 1.

slide-46
SLIDE 46

Part II - The extenders Fα,n (3/3)

⊳ and Fα,n Fα′,n′ ⊳ Fα,n iff α′ < α and n′ ≥ n.

slide-47
SLIDE 47

Part II - The extenders Fα,n (3/3)

⊳ and Fα,n Fα′,n′ ⊳ Fα,n iff α′ < α and n′ ≥ n.

◮ We want to replace the extenders Fα,n with normal measure

Uα,n preserving the ⊳ structure.

slide-48
SLIDE 48

Part II - The extenders Fα,n (3/3)

⊳ and Fα,n Fα′,n′ ⊳ Fα,n iff α′ < α and n′ ≥ n.

◮ We want to replace the extenders Fα,n with normal measure

Uα,n preserving the ⊳ structure.

◮ We force over V to collapse the generators of the extenders

Fα,n.

slide-49
SLIDE 49

Part II - The extenders Fα,n (3/3)

⊳ and Fα,n Fα′,n′ ⊳ Fα,n iff α′ < α and n′ ≥ n.

◮ We want to replace the extenders Fα,n with normal measure

Uα,n preserving the ⊳ structure.

◮ We force over V to collapse the generators of the extenders

Fα,n.

◮ We want to do this carefully and avoid introducing “too

many” new normal measures.

slide-50
SLIDE 50

Part II - from Fα,n to Uα,n (1/3)

Force with P = Pν, ˙ Qν | ν ≤ κ. Friedman-Magidor (nonstationary) support iteration of Collapsing and Coding posets: 1. ˙ Qν is not trivial iff ν ≤ κ is an inaccessible limit of measurable cardinals 2. ˙ Qν = Coll(ν+, θ(ν)++) ∗ Code(ν+, gν) where

◮ Coll(ν+, θ(ν)++) introduces a surjection gν : ν+ → θ(ν)++

slide-51
SLIDE 51

Part II - from Fα,n to Uα,n (1/3)

Force with P = Pν, ˙ Qν | ν ≤ κ. Friedman-Magidor (nonstationary) support iteration of Collapsing and Coding posets: 1. ˙ Qν is not trivial iff ν ≤ κ is an inaccessible limit of measurable cardinals 2. ˙ Qν = Coll(ν+, θ(ν)++) ∗ Code(ν+, gν) where

◮ Coll(ν+, θ(ν)++) introduces a surjection gν : ν+ → θ(ν)++ ◮ Code(ν+, gν) introduces a club Cν ⊂ ν+.

Cν codes gν and itself by destroying certain stationary sets from a pre chosen sequence Ti | i < ν+

  • 3. Let G ⊂ P be a V −generic filter
slide-52
SLIDE 52

Part II - from Fα,n to Uα,n (2/3)

The Friedman-Magidor iteration style guarantees that jα,n : V → Mα,n = Ult(V , Fα,n) uniquely extends to j∗

α,n : V [G] → M∗ α,n = Mα,n[Gα,n], where ◮ V [G] and Mα,n[Gα,n] agree on the collapsing generic function

gκ : κ+ → θ++ forced at stage κ.

slide-53
SLIDE 53

Part II - from Fα,n to Uα,n (2/3)

The Friedman-Magidor iteration style guarantees that jα,n : V → Mα,n = Ult(V , Fα,n) uniquely extends to j∗

α,n : V [G] → M∗ α,n = Mα,n[Gα,n], where ◮ V [G] and Mα,n[Gα,n] agree on the collapsing generic function

gκ : κ+ → θ++ forced at stage κ.

◮ Every ordinal γ < θ++ as j∗ α,n(f )(κ) for some f ∈ κκ in V [G].

slide-54
SLIDE 54

Part II - from Fα,n to Uα,n (3/3)

It follows that j∗

α,n : V [G] → M∗ α,n ∼

= Ult(V [G], Uα,n) where Uα,n = {X ⊆ κ | κ ∈ j∗

α,n(X)}

slide-55
SLIDE 55

Part II - from Fα,n to Uα,n (3/3)

It follows that j∗

α,n : V [G] → M∗ α,n ∼

= Ult(V [G], Uα,n) where Uα,n = {X ⊆ κ | κ ∈ j∗

α,n(X)}

  • 1. Uα′,n′ ⊳ Uα,n iff α′ < α and n′ ≥ n.
  • 2. Uα,n, n < ω, α < λ, are the only normal measures on κ

in V [G].

slide-56
SLIDE 56

Part II - from Fα,n to Uα,n (3/3)

It follows that j∗

α,n : V [G] → M∗ α,n ∼

= Ult(V [G], Uα,n) where Uα,n = {X ⊆ κ | κ ∈ j∗

α,n(X)}

  • 1. Uα′,n′ ⊳ Uα,n iff α′ < α and n′ ≥ n.
  • 2. Uα,n, n < ω, α < λ, are the only normal measures on κ

in V [G]. Next: we use ⊳(κ) in V [G] to realize non-tame orders.

slide-57
SLIDE 57

Part II - First Application - Sω,2

Suppose that F = F0, F1, λ = 2 The normal measures on κ in V [G] are U0,n, U1,n, n < ω, and ⊳(κ) = {(U0,n′, U1,n) | n′ ≥ n}.

slide-58
SLIDE 58

Part II - First Application - Sω,2

Suppose that F = F0, F1, λ = 2 The normal measures on κ in V [G] are U0,n, U1,n, n < ω, and ⊳(κ) = {(U0,n′, U1,n) | n′ ≥ n}.

  • U0,0
  • U0,1
  • U0,2

. . . . . . . . .

  • U0,n

. . . . . .

  • U1,0
  • U1,1
  • U1,2. . . . . . . . .
  • U1,n. . . . . .
slide-59
SLIDE 59

Part II - Second Application - R2,2

Suppose that F = F0, F1, F2, λ = 3. In V [G] let S = {U0,0, U1,0, U1,1, U2,1}. ⊳(κ) ↾ S ∼ = R2,2.

  • U0,0
  • U1,0
  • U1,1
  • U2,1
slide-60
SLIDE 60

Part II - Second Application - R2,2

Suppose that F = F0, F1, F2, λ = 3. In V [G] let S = {U0,0, U1,0, U1,1, U2,1}. ⊳(κ) ↾ S ∼ = R2,2.

  • U0,0
  • U1,0
  • U1,1
  • U2,1

◮ Separation by Sets: There is X ⊂ κ so that the

X ∈ U ⇐ ⇒ U ∈ S.

slide-61
SLIDE 61

Part II - Second Application - R2,2

Suppose that F = F0, F1, F2, λ = 3. In V [G] let S = {U0,0, U1,0, U1,1, U2,1}. ⊳(κ) ↾ S ∼ = R2,2.

  • U0,0
  • U1,0
  • U1,1
  • U2,1

◮ Separation by Sets: There is X ⊂ κ so that the

X ∈ U ⇐ ⇒ U ∈ S.

◮ The final cut forcing by X, PX = PX ν , QX ν | ν ∈ X ∪ {κ} is

a variant of the Friedman-Magidor forcing where QX

ν = Code(ν+, ∅), ν ∈ X ∪ {κ}. ◮ The measures U ∈ S are the only measures which extend in

V [G]PX .

◮ In the final cut generic extension, ⊳(κ) ∼

= R2,2.

slide-62
SLIDE 62

Part II - Third Application

Suppose F = Fk | k < ω. In V [G] define blocks Bn, n < ω: Bn = {Ui,n | kn ≤ i ≤ kn + n}, kn = n(n + 1) 2

1 2

  • U0,0

B0 U1,1

  • U2,1
  • B1

U3,2

  • U4,2
  • U5,2
  • B2

n ... .........

  • Ukn,n
  • Ukn+1,n

. . . . . . . . . . . .

  • Ukn+n, n

Bn

Let B =

n<ω Bn. There is a final cut extension V ∗ where

⊳(κ)V ∗ ∼ = ⊳(κ)V [G] ↾ B

slide-63
SLIDE 63

Part II - Realizing Arbitrary Well-Founded Orders

To realize arbitrary well founded ordered we use auxiliary orders: Auxiliary orders R∗

λ,ρ

For an ordinal λ and a cardinal ρ,

  • 1. R∗

λ,ρ = λ × ρ2

  • 2. (α′, c′) <R∗

β,ρ (α, c) if and only if ◮ α′ < α, and ◮ c′ ≥ c (pointwise)

slide-64
SLIDE 64

Part II - Realizing Arbitrary Well-Founded Orders

To realize arbitrary well founded ordered we use auxiliary orders: Auxiliary orders R∗

λ,ρ

For an ordinal λ and a cardinal ρ,

  • 1. R∗

λ,ρ = λ × ρ2

  • 2. (α′, c′) <R∗

β,ρ (α, c) if and only if ◮ α′ < α, and ◮ c′ ≥ c (pointwise)

(S, <S) embeds into R∗

rank(S,<S),|S|

slide-65
SLIDE 65

Part II - Revised Ground Model Assumptions

Suppose we want to realize (S, <S). May assume that S ⊂ R∗

λ,ρ,

λ < κ+, ρ ≤ κ.

slide-66
SLIDE 66

Part II - Revised Ground Model Assumptions

Suppose we want to realize (S, <S). May assume that S ⊂ R∗

λ,ρ,

λ < κ+, ρ ≤ κ. Previous Construction Revised Construction κ < θ κ < θ = θi | i < ρ θ = supi<ρ θ+

i

  • F = Fα | α < λ

(κ, θ++)-extenders (θ + 2)-strong

  • F = Fα | α < λ

(κ, θ+)-extenders (θ + 1)-strong in: n−th iterated ultrapower by Uθ ic, c ∈ ρ2: iterated ultrapower by the Uθi s.t. c(i) = 1 Fα,n = in(Fα) Fα,c = ic(Fα) Fα′,n′ ⊳ Fα,n ⇐ ⇒ α′ < α and n′ ≥ n Fα′,c′ ⊳ Fα,c ⇐ ⇒ α′ < α and c′ ≥ c ⇐ ⇒ (α′, c′) <R∗ (α, c)

slide-67
SLIDE 67

Regaining Completeness for Fα,c

A Problem: If |c−1(1)| ≥ ℵ0 then θi | c(i) = 1∈ Ult(V , Fα,c)

slide-68
SLIDE 68

Regaining Completeness for Fα,c

A Problem: If |c−1(1)| ≥ ℵ0 then θi | c(i) = 1∈ Ult(V , Fα,c)

◮ To fix this, we force with a Magidor iteration of one-point

Prikry forcing P1 = P1

µ, ˙

Q1

µ | µ < κ.

slide-69
SLIDE 69

Regaining Completeness for Fα,c

A Problem: If |c−1(1)| ≥ ℵ0 then θi | c(i) = 1∈ Ult(V , Fα,c)

◮ To fix this, we force with a Magidor iteration of one-point

Prikry forcing P1 = P1

µ, ˙

Q1

µ | µ < κ. ◮

˙ Q1

µ = Q(Uµ) is the one-point Prikry forcing, choosing a single

Prikry point d(µ) < µ

slide-70
SLIDE 70

Regaining Completeness for Fα,c

A Problem: If |c−1(1)| ≥ ℵ0 then θi | c(i) = 1∈ Ult(V , Fα,c)

◮ To fix this, we force with a Magidor iteration of one-point

Prikry forcing P1 = P1

µ, ˙

Q1

µ | µ < κ. ◮

˙ Q1

µ = Q(Uµ) is the one-point Prikry forcing, choosing a single

Prikry point d(µ) < µ

◮ Qµ is nontrivial when µ = θi(ν) for ν is inaccessible limit of

measurable cardinals and c(i) = 1.

slide-71
SLIDE 71

Regaining Completeness for Fα,c

A Problem: If |c−1(1)| ≥ ℵ0 then θi | c(i) = 1∈ Ult(V , Fα,c)

◮ To fix this, we force with a Magidor iteration of one-point

Prikry forcing P1 = P1

µ, ˙

Q1

µ | µ < κ. ◮

˙ Q1

µ = Q(Uµ) is the one-point Prikry forcing, choosing a single

Prikry point d(µ) < µ

◮ Qµ is nontrivial when µ = θi(ν) for ν is inaccessible limit of

measurable cardinals and c(i) = 1.

jα,c : V → Mα,c = Ult(V , Fα,c) extends to j1

α,c : V [G 1] → Mα,c[G 1 α,c] ∋ θi | c(i) = 1

slide-72
SLIDE 72

Regaining Completeness for Fα,c

A Problem: If |c−1(1)| ≥ ℵ0 then θi | c(i) = 1∈ Ult(V , Fα,c)

◮ To fix this, we force with a Magidor iteration of one-point

Prikry forcing P1 = P1

µ, ˙

Q1

µ | µ < κ. ◮

˙ Q1

µ = Q(Uµ) is the one-point Prikry forcing, choosing a single

Prikry point d(µ) < µ

◮ Qµ is nontrivial when µ = θi(ν) for ν is inaccessible limit of

measurable cardinals and c(i) = 1.

jα,c : V → Mα,c = Ult(V , Fα,c) extends to j1

α,c : V [G 1] → Mα,c[G 1 α,c] ∋ θi | c(i) = 1 ◮ The (κ, θ+)−extender F 1 α,c derived from j1 α,c, is κ−complete

slide-73
SLIDE 73

Regaining Completeness for Fα,c

A Problem: If |c−1(1)| ≥ ℵ0 then θi | c(i) = 1∈ Ult(V , Fα,c)

◮ To fix this, we force with a Magidor iteration of one-point

Prikry forcing P1 = P1

µ, ˙

Q1

µ | µ < κ. ◮

˙ Q1

µ = Q(Uµ) is the one-point Prikry forcing, choosing a single

Prikry point d(µ) < µ

◮ Qµ is nontrivial when µ = θi(ν) for ν is inaccessible limit of

measurable cardinals and c(i) = 1.

jα,c : V → Mα,c = Ult(V , Fα,c) extends to j1

α,c : V [G 1] → Mα,c[G 1 α,c] ∋ θi | c(i) = 1 ◮ The (κ, θ+)−extender F 1 α,c derived from j1 α,c, is κ−complete ◮ We can now collapse the generators of F 1 α,c as before, and use

the induced normal measures Uα,c to realize (S, <S)

slide-74
SLIDE 74

Theorem 2 (BN) Let V = L[E] be a core model. Suppose that κ is a cardinal in V and (S, <S) is a well-founded order of cardinality ≤ κ, so that

  • 1. there are |S| measurable cardinals above κ; let θ be the

supremum of the their successors,

  • 2. there is a ⊳−increasing sequence of (θ + 1)− strong

extenders F = Fα | α < rank(S, <S) Then there is a generic extension V ∗ of V in which ⊳(κ)V ∗ ∼ = (S, <S).

slide-75
SLIDE 75

Theorem 2 (BN) Let V = L[E] be a core model. Suppose that κ is a cardinal in V and (S, <S) is a well-founded order of cardinality ≤ κ, so that

  • 1. there are |S| measurable cardinals above κ; let θ be the

supremum of the their successors,

  • 2. there is a ⊳−increasing sequence of (θ + 1)− strong

extenders F = Fα | α < rank(S, <S) Then there is a generic extension V ∗ of V in which ⊳(κ)V ∗ ∼ = (S, <S). Corollary (sufficient large cardinal assumptions): There is a class forcing extension in which every well-founded order (S, <S) is isomorphic to ⊳(κ) at some κ.

slide-76
SLIDE 76

Theorem 2 (BN) Let V = L[E] be a core model. Suppose that κ is a cardinal in V and (S, <S) is a well-founded order of cardinality ≤ κ, so that

  • 1. there are |S| measurable cardinals above κ; let θ be the

supremum of the their successors,

  • 2. there is a ⊳−increasing sequence of (θ + 1)− strong

extenders F = Fα | α < rank(S, <S) Then there is a generic extension V ∗ of V in which ⊳(κ)V ∗ ∼ = (S, <S). Corollary (sufficient large cardinal assumptions): There is a class forcing extension in which every well-founded order (S, <S) is isomorphic to ⊳(κ) at some κ.

slide-77
SLIDE 77

Theorem 2 (BN) Let V = L[E] be a core model. Suppose that κ is a cardinal in V and (S, <S) is a well-founded order of cardinality ≤ κ, so that

  • 1. there are |S| measurable cardinals above κ; let θ be the

supremum of the their successors,

  • 2. there is a ⊳−increasing sequence of (θ + 1)− strong

extenders F = Fα | α < rank(S, <S) Then there is a generic extension V ∗ of V in which ⊳(κ)V ∗ ∼ = (S, <S). Corollary (sufficient large cardinal assumptions): There is a class forcing extension in which every well-founded order (S, <S) is isomorphic to ⊳(κ) at some κ.

slide-78
SLIDE 78

Questions

  • 1. Realizing non-tame orders: What is the large cardinal

assumption required to realize the non-tame orders R2,2 Sω,2? Are overlapping extenders necessary?

slide-79
SLIDE 79

Questions

  • 1. Realizing non-tame orders: What is the large cardinal

assumption required to realize the non-tame orders R2,2 Sω,2? Are overlapping extenders necessary?

  • 2. Gaps realizing tame orders: What is the large cardinal

assumption required to realize S2,2?

  • rank(S2,2) = 2 but Trank(S2,2) = 3. Can we realize S2,2 from
  • (κ) = 2?
slide-80
SLIDE 80

Questions

  • 1. Realizing non-tame orders: What is the large cardinal

assumption required to realize the non-tame orders R2,2 Sω,2? Are overlapping extenders necessary?

  • 2. Gaps realizing tame orders: What is the large cardinal

assumption required to realize S2,2?

  • rank(S2,2) = 2 but Trank(S2,2) = 3. Can we realize S2,2 from
  • (κ) = 2?
  • 3. Is it possible to realize arbitrary well-founded orders of size κ+

as ⊳(κ) ?

slide-81
SLIDE 81

Thank You!