In mathematical logic and set theory, an ordinal collapsing function (or projection function) is a technique for defining (notations for) certain recursive large countable ordinals, whose principle is to give names to certain ordinals much larger than the one being defined, perhaps even large cardinals (though they can be replaced with recursively large ordinals at the cost of extra technical difficulty), and then "collapse" them down to a system of notations for the sought-after ordinal. For this reason, ordinal collapsing functions are described as an impredicative manner of naming ordinals.
The details of the definition of ordinal collapsing functions vary, and get more complicated as greater ordinals are being defined, but the typical idea is that whenever the notation system "runs out of fuel" and cannot name a certain ordinal, a much larger ordinal is brought "from above" to give a name to that critical point. An example of how this works will be detailed below, for an ordinal collapsing function defining the Bachmann–Howard ordinal (i.e., defining a system of notations up to the Bachmann–Howard ordinal).
The use and definition of ordinal collapsing functions is inextricably intertwined with the theory of ordinal analysis, since the large countable ordinals defined and denoted by a given collapse are used to describe the ordinal-theoretic strength of certain formal systems, typically[1] [2] subsystems of analysis (such as those seen in the light of reverse mathematics), extensions of Kripke–Platek set theory, Bishop-style systems of constructive mathematics or Martin-Löf-style systems of intuitionistic type theory.
Ordinal collapsing functions are typically denoted using some variation of either the Greek letter
\psi
\theta
The choice of the ordinal collapsing function given as example below imitates greatly the system introduced by Buchholz[3] but is limited to collapsing one cardinal for clarity of exposition. More on the relation between this example and Buchholz's system will be said below.
Let
\Omega
\omega1
\varepsilon
\omega1
We define a function
\psi
\alpha
\psi(\alpha)
\alpha
Assume
\psi(\beta)
\beta<\alpha
\psi(\alpha)
Let
C(\alpha)
0
1
\omega
\Omega
\psi\upharpoonright\alpha
\psi
\beta<\alpha
C(\alpha)0=\{0,1,\omega,\Omega\}
C(\alpha)n+1=C(\alpha)n\cup\{\beta1+\beta2,\beta1\beta2,{\beta
\beta2 | |
1} |
:\beta1,\beta2\inC(\alpha)n\}\cup\{\psi(\beta):\beta\inC(\alpha)n\land\beta<\alpha\}
n
C(\alpha)
C(\alpha)n
n
Then
\psi(\alpha)
C(\alpha)
In a more concise (although more obscure) way:
\psi(\alpha)
0
1
\omega
\Omega
\psi
\alpha
Here is an attempt to explain the motivation for the definition of
\psi
\Omega
\psi
To clarify how the function
\psi
First consider
C(0)
0,1,2,3,\omega,\omega+1,\omega+2,\omega ⋅ 2,\omega ⋅ 3,\omega2,\omega3,\omega\omega,
\omega\omega | |
\omega |
\Omega,\Omega+1,\omega\Omega+1,\Omega\Omega
\varepsilon0
\omega
\omega\omega
\omega\omega | |
\omega |
\Omega
\varepsilon\Omega+1
\Omega
\Omega\Omega
\Omega\Omega | |
\Omega |
\psi(0)=\varepsilon0
Similarly,
C(1)
0
1
\omega
\Omega
\varepsilon0
\varepsilon1
\psi(1)=\varepsilon1
\psi(\alpha)=\varepsilon\alpha
\alpha
\alpha<\varepsilon\alpha
\psi(\alpha)=\varepsilon\alpha=\phi1(\alpha)
\alpha\leq\zeta0
\zeta0=\phi2(0)
\alpha\mapsto\varepsilon\alpha
(Here, the
\phi
\phi1(\alpha)=\varepsilon\alpha
Now
\psi(\zeta0)=\zeta0
\psi(\zeta0+1)
\zeta0
\phi1\colon\alpha\mapsto\varepsilon\alpha
C(\alpha)
\alpha\leq\Omega
\psi
\zeta0
\psi(\alpha)=\zeta0
\zeta0\leq\alpha\leq\Omega
Again,
\psi(\Omega)=\zeta0
\psi(\Omega+1)
\Omega
C(\alpha)
\psi(\Omega)=\zeta0
C(\Omega+1)
0
1
\omega
\Omega
\phi1\colon\alpha\mapsto\varepsilon\alpha
\zeta0
\zeta0
C(\Omega+1)
\varepsilon | |
\zeta0+1 |
\varepsilon
\zeta0
We say that the definition
\psi(\Omega)=\zeta0
\psi
\psi(\Omega+1)=
\varepsilon | |
\zeta0+1 |
\Omega
\zeta0
The fact that
\psi(\Omega+\alpha)
\varepsilon | |
\zeta0+\alpha |
\alpha\leq\zeta1=\phi2(1)
\psi(\Omega+\zeta0)=
\varepsilon | |
\zeta0 ⋅ 2 |
\zeta0
\zeta1=\phi2(1)
\alpha\mapsto\varepsilon\alpha
\zeta0
\zeta1
\zeta0
\varepsilon
\psi(\Omega ⋅ 2)=\zeta1
The same reasoning shows that
\psi(\Omega(1+\alpha))=\phi2(\alpha)
\alpha\leq\phi3(0)=η0
\phi2
\phi1\colon\alpha\mapsto\varepsilon\alpha
\phi3(0)
\phi2
\psi(\Omega2)=\phi3(0)
Again, we can see that
\psi(\Omega\alpha)=\phi1+\alpha(0)
\Gamma0
\alpha\mapsto\phi\alpha(0)
\psi(\Omega\Omega)=\Gamma0
We have
\psi(\Omega\Omega+\Omega\alpha)=
\phi | |
\Gamma0+\alpha |
(0)
\alpha\leq\Gamma1
\Gamma1
\alpha\mapsto\phi\alpha(0)
\alpha\mapsto\Gamma\alpha
\phi(1,0,\alpha)
\psi(\Omega\Omega(1+\alpha))=\Gamma\alpha
\phi(1,1,0)
\alpha\mapsto\Gamma\alpha
\psi(\Omega\Omega+1)
\phi(2,0,0)
\alpha\mapsto\phi(1,\alpha,0)
\psi(\Omega\Omega ⋅ )
\Omega2 | |
\psi(\Omega |
)
\phi(\alpha,\beta,\gamma)
\Omega\omega | |
\psi(\Omega |
)
\phi( ⋅ )
\Omega\Omega | |
\psi(\Omega |
)
\phi( ⋅ )
\psi(\varepsilon\Omega+1)
\psi(\Omega)
\psi(\Omega\Omega)
\Omega\Omega | |
\psi(\Omega |
)
\psi
We now explain more systematically how the
\psi
Recall that if
\delta
\omega
\omega
\varepsilon0
\Omega
\alpha
\beta1 | |
\delta |
\gamma1+\ldots+
\betak | |
\delta |
\gammak
k
\gamma1,\ldots,\gammak
\delta
\beta1>\beta2> … >\betak
\betak=0
\delta
\delta=\omega
\alpha=\delta\alpha
\betai
\alpha
\alpha<\delta
k\leq1
\gamma1=\alpha
If
\alpha
\varepsilon\Omega+1
\Omega
\gammai<\Omega
\betai<\alpha
\alpha<\varepsilon\Omega+1
\Omega
\Omega
\alpha
<\Omega
\Omega
\alpha
\psi
\psi(\alpha)=\psi(\beta)
\beta<\alpha
C(\alpha)=C(\beta)
\beta'
\beta\leq\beta'<\alpha
C(\alpha)
\psi
\psi(\alpha)
C(\alpha)
C(\beta)
C(\alpha)
\gamma=\psi(\alpha)
\psi
\varepsilon
\beta\mapsto\omega\beta
C(\alpha)
C(\alpha)
\delta
\varepsilon
\alpha
\psi(\beta)<\delta
\beta<\alpha
\Omega
C(\alpha)
\delta
C'
\Omega
\delta
C'
\delta
\varepsilon
C'
\psi(\beta)
\beta<\alpha
0
1
\omega
\Omega
C'\supseteqC(\alpha)
\psi(\alpha)\leq\delta
\delta\not\inC(\alpha)
\varepsilon
\psi
\psi
\psi
\varepsilon
\delta
\varepsilon
\psi
\alpha
\beta
\psi(\beta)<\delta
\psi(\alpha)\leq\delta
\psi(\alpha)<\delta
\alpha
\psi(\alpha)=\delta
\psi(\alpha)=\delta
C(\alpha)
\gamma
\varepsilon\Omega+1
\Omega
\delta
\delta
\varepsilon\Omega+1
\Omega
\delta
C(\alpha)
\psi(\beta)<\delta
\beta<\alpha
\alpha
\psi(\alpha)=\delta
\psi(\alpha)=\psi(\beta)
\beta<\alpha
C(\alpha)=C(\beta)
\alpha
\psi(\alpha)=\delta
Using the facts above, we can define a (canonical) ordinal notation for every
\gamma
\gamma
If
\gamma
\varepsilon0
\gamma
\varepsilon
\delta
\gamma
\varepsilon
\delta<\gamma
\delta
\delta
\gamma
\gamma
It remains to deal with the case where
\gamma=\delta
\varepsilon
\delta=\psi(\alpha)
\alpha<\varepsilon\Omega+1
\alpha
\psi
\Omega
\alpha
\delta
C(\alpha)
\alpha
C(\alpha+1)=C(\alpha)
\psi
\alpha
\psi(\alpha+1)=\psi(\alpha)=\delta
\alpha
Note: Actually, we have defined canonical notations not just for ordinals below the Bachmann–Howard ordinal but also for certain uncountable ordinals, namely those whose
\Omega
\Omega
\psi
For ordinals less than
\varepsilon0=\psi(0)
For ordinals less than
\varepsilon1=\psi(1)
\varepsilon0
| |||||
\omega |
\omega\omega | |
{\varepsilon | |
0} |
\omega\omega | |
\psi(0) |
\varepsilon2=\psi(2)
\varepsilon1
\varepsilon0
| |||||
\omega |
\varepsilon0\omega | |
{\varepsilon | |
1} |
\psi(1)\psi(0)\omega
\zeta0=\psi(\Omega)
\varepsilon
Beyond this, we may need to express ordinals beyond
\Omega
\Omega
\varepsilon
Note that while
\psi(\varepsilon\Omega+1)
The notations thus defined have the property that whenever they nest
\psi
\psi
\Omega
\alpha
\alpha
\psi(\alpha)=\delta
\varepsilon
\delta
\delta
\psi(\psi(\Omega)+1)
\psi(\Omega)=\zeta0
\psi
\zeta0
\Omega
Canonicalness can be checked recursively: an expression is canonical if and only if it is either the iterated Cantor normal form of an ordinal less than
\varepsilon0
\delta
\delta=\psi(\alpha)
\alpha
\Omega
\delta
\Omega
\psi
\psi
For example,
\psi(\Omega\omega+1\psi(\Omega)+\psi(\Omega\omega)
\psi(\Omega2) | |
42)\psi(1729)\omega
\varphi1(\varphi\omega+1(\varphi2(0))+
\varphi3(0) | |
\varphi | |
\omega(0) |
\varphi1(1729)\omega | |
42) |
Concerning the order, one might point out that
\psi(\Omega\Omega)
\psi(\Omega\psi(\Omega))=
\varphi | |
\varphi2(0) |
(0)
\Omega
\psi
\psi(\Omega\psi(\Omega))=
\varphi | |
\varphi2(0) |
(0)
\psi(\Omega)\psi(\Omega)=
\varphi2(0) | |
\varphi | |
2(0) |
\Omega\psi(\Omega)
\Omega
\psi(\Omega)
\psi(\Omega\Omega)
\psi(\Omega)\psi(\Omega)
\psi(\Omega+1)
To witness the fact that we have defined notations for ordinals below the Bachmann–Howard ordinal (which are all of countable cofinality), we might define standard sequences converging to any one of them (provided it is a limit ordinal, of course). Actually we will define canonical sequences for certain uncountable ordinals, too, namely the uncountable ordinals of countable cofinality (if we are to hope to define a sequence converging to them...) which are representable (that is, all of whose
\Omega
The following rules are more or less obvious, except for the last:
\delta
\alpha=
\beta1 | |
\delta |
\gamma1+ … +
\betak | |
\delta |
\gammak
\delta
\omega
\psi( … )
\Omega
k
\alpha=0
\betak
\gammak
\alpha
\gammak
\gammak
\gammak
\gammak
\betak
\betak | |
\delta |
\gammak
\betak | |
\delta |
(\gammak-1)+
\betak | |
\delta |
\betak
\gammak
\betak
\betak | |
\delta |
\gammak
\betak | |
\delta |
(\gammak-1)+
\betak-1 | |
\delta |
\delta
\delta
\delta
\omega
0,1,2,3,\ldots
\delta
\delta=\psi(0)
\delta
\omega,\omega\omega,
\omega\omega | |
\omega |
,\ldots
\delta=\psi(\alpha+1)
\delta
\psi(\alpha),\psi(\alpha)\psi(\alpha),
\psi(\alpha)\psi(\alpha) | |
\psi(\alpha) |
,\ldots
\delta=\psi(\alpha)
\alpha
\delta
\psi
\alpha
\psi
\delta=\psi(\alpha)
\alpha
\Omega
\alpha
\rho<\alpha
\psi
\rho
\alpha
\rho
\xi\mapstoh(\psi(\xi))
\Omega
\alpha
\alpha
\Omega
\Omega
\alpha=h(\Omega)
\psi(\xi)
\xi
0
\xi\mapstoh(\psi(\xi))
0,h(\psi(0)),h(\psi(h(\psi(0)))),\ldots
\rho
\psi(\alpha)=\psi(\rho)
\psi(0)
\psi(h(\psi(0)))
\psi(h(\psi(h(\psi(0)))))
n
0
\delta
\delta[n]
\delta[0]=\psi(0)
\delta[n]=\psi(h(\delta[n-1]))
Here are some examples for the last (and most interesting) case:
\psi(\Omega)
\psi(0)
\psi(\psi(0))
\psi(\psi(\psi(0)))
\rho=\psi(\Omega)=\zeta0
\psi
\Omega
\psi(\Omega2)
\psi(0)
\psi(\Omega+\psi(0))
\psi(\Omega+\psi(\Omega+\psi(0))),\ldots
\psi
\rho=\Omega+\psi(\Omega2)=\Omega+\zeta1
\psi
\Omega2
\psi(\Omega2)
\psi(0),\psi(\Omega\psi(0)),\psi(\Omega\psi(\Omega\psi(0))),\ldots
\psi
\rho=\Omega\psi(\Omega2)
\psi(\Omega23+\Omega)
\psi(0),\psi(\Omega23+\psi(0)),\psi(\Omega23+\psi(\Omega23+\psi(0))),\ldots
\psi
\rho=\Omega23+\psi(\Omega23+\Omega)
\psi(\Omega\Omega)
\psi(0),\psi(\Omega\psi(0)),
\psi(\Omega\psi(0)) | |
\psi(\Omega |
),\ldots
\psi
\rho=
\psi(\Omega\Omega) | |
\Omega |
\psi(\Omega\Omega3)
\psi(0),\psi(\Omega\Omega2+\Omega\psi(0)),\psi(\Omega\Omega
\psi(\Omega\Omega2+\Omega\psi(0)) | |
2+\Omega |
),\ldots
\psi
\rho=\Omega\Omega2+
\psi(\Omega\Omega3) | |
\Omega |
\psi(\Omega\Omega+1)
\psi(0),\psi(\Omega\Omega\psi(0)),\psi(\Omega\Omega\psi(\Omega\Omega\psi(0))),\ldots
\psi
\rho=\Omega\Omega\psi(\Omega\Omega+1)
\Omega2+\Omega3 | |
\psi(\Omega |
)
\psi(0),
\Omega2+\Omega2+\psi(0) | |
\psi(\Omega |
),
| |||||||||||
\psi(\Omega |
),\ldots
Here are some examples of the other cases:
\omega2
0
\omega
\omega2
\omega3
\psi(\omega\omega)
\psi(1)
\psi(\omega)
\psi(\omega2)
\psi(\omega3)
\psi(\Omega)\omega
1
\psi(\Omega)
\psi(\Omega)2
\psi(\Omega)3
\psi(\Omega+1)
\psi(\Omega)
\psi(\Omega)\psi(\Omega)
\psi(\Omega)\psi(\Omega) | |
\psi(\Omega) |
\psi(\Omega+\omega)
\psi(\Omega)
\psi(\Omega+1)
\psi(\Omega+2)
\psi(\Omega+3)
\psi(\Omega\omega)
\psi(0)
\psi(\Omega)
\psi(\Omega2)
\psi(\Omega3)
\psi(\Omega\omega)
\psi(1)
\psi(\Omega)
\psi(\Omega2)
\psi(\Omega3)
\psi(\Omega\psi(0))
\psi(\Omega\omega)
\omega\omega | |
\psi(\Omega |
)
| |||||
\psi(\Omega |
)
\psi(0)
\psi(\Omega\psi(\Omega))
\psi(\Omega\psi(0))
\psi(\Omega\psi(\psi(0)))
\psi(\Omega\psi(\psi(\psi(0))))
\psi(\Omega)
Even though the Bachmann–Howard ordinal
\psi(\varepsilon\Omega+1)
\psi(\Omega)
\psi(\Omega\Omega)
\Omega\Omega | |
\psi(\Omega |
)
Start with any ordinal less than or equal to the Bachmann–Howard ordinal, and repeat the following process so long as it is not zero:
Then it is true that this process always terminates (as any decreasing sequence of ordinals is finite); however, like (but even more so than for) the hydra game:
To give some flavor of what the process feels like, here are some steps of it: starting from
\Omega\omega | |
\psi(\Omega |
)
\Omega3 | |
\psi(\Omega |
)
\Omega2\psi(0) | |
\psi(\Omega |
)
\Omega2\omega\omega | |
\psi(\Omega |
)
\Omega2\omega3 | |
\psi(\Omega |
)
\Omega2\omega23 | |
\psi(\Omega |
)
\Omega2(\omega22+\omega) | |
\psi(\Omega |
)
\Omega2(\omega22+1) | |
\psi(\Omega |
)
| |||||||||||
\psi(\Omega |
)
| |||||||||||||||||||||
\psi(\Omega |
)
Concerning the first statement, one could introduce, for any ordinal
\alpha
\psi(\varepsilon\Omega+1)
f\alpha(n)
n
f\alpha(n)=f\alpha[n](n)+1
f\alpha
f | |
\omega\omega |
(n)
nn
f | |
\psi(\Omega\omega) |
(n)
A(n,n)
f | |
\psi(\varepsilon\Omega+1) |
(n)
g\alpha(n)=g\alpha[n](n+1)+1
g\psi(0)(n)
g | |||||||
|
(n)
Concerning the second statement, a precise version is given by ordinal analysis: for example, Kripke–Platek set theory can prove[4] that the process terminates for any given
\alpha
\varepsilon0
It is instructive (although not exactly useful) to make
\psi
If we alter the definition of
\psi
C(\alpha)
\psi(0)=\omega\omega
0
1
\omega
\psi(1)=
\omega2 | |
\omega |
\psi(\omega)=
\omega\omega | |
\omega |
\psi(\psi(0))=
| |||||
\omega |
\psi(\Omega)=\varepsilon0
\psi(\Omega+1)=
\omega | |
{\varepsilon | |
0} |
\psi(\Omega2)=\varepsilon1
\Omega
\psi(\Omega2)=\varphi2(0)
\psi(\Omega3)=\varphi3(0)
\Omega\omega
\psi(\Omega\omega)=\phi\omega(0)
\psi(\Omega\omega)
If we alter the definition of
\psi
\psi(0)=\omega2
\psi(1)=\omega3
\psi(\psi(0))=
\omega2 | |
\omega |
\psi(\Omega)=\varepsilon0
\psi(\Omega+1)=\varepsilon0\omega
\psi(\Omega2)=\varepsilon1
\psi(\Omega3)=\varepsilon2
\Omega
\psi(\Omega\omega)=\varepsilon\omega=\varphi1(\omega)
If we alter the definition even more, to allow nothing except psi, we get
\psi(0)=1
\psi(\psi(0))=2
\psi(\omega)=\omega+1
\psi(\psi(\omega))=\omega+2
\psi(\Omega)=\omega2
\Omega
\omega2
In both cases, we find that the limitation on the weakened
\psi
We know that
\psi(\varepsilon\Omega+1)
\psi(\varepsilon\Omega+1+1)
\varepsilon\Omega+1
C(\alpha)
\alpha
\varepsilon
\psi
\psi(\Omega+1)
\varepsilon | |
\varphi2(0)+1 |
\varepsilon\Omega+1
Let
\psi1(\alpha)
\Omega2
\psi1
\alpha
Here,
\Omega2
\psi1
\Omega=\omega1
\Omega2=\omega2
For example,
\psi1(0)=\Omega
\psi1(\alpha)=\varepsilon\Omega+\alpha
\psi1(\Omega)=\psi1(\psi1(0))=\varepsilon\Omega
\psi1(\psi1(1))=
\varepsilon | |
\varepsilon\Omega+1 |
\zeta\Omega+1
\xi\mapsto\varepsilon\xi
\Omega
\psi1(0)
\psi1(\psi1(0))
\psi1(\alpha)=\zeta\Omega+1
\Omega2
\psi(\Omega)
\psi1(\Omega2)=\zeta\Omega+1
\psi1(\Omega2+1)=
\varepsilon | |
\zeta\Omega+1+1 |
The
\psi1
\psi1(\varepsilon
\Omega2+1 |
)
\psi1(\Omega2)
\psi1({\Omega
\Omega2 | |
2} |
)
Now we can reinject these notations in the original
\psi
\psi(\alpha)
0
1
\omega
\Omega
\Omega2
\psi1
\psi
\alpha
This modified function
\psi
\psi(\psi1(1))
\psi(\psi1(1)+1)
\varepsilon | |
\psi(\psi1(1))+1 |
\varepsilon
\Omega
\Omega2
\Omega2
A variation on this scheme, which makes little difference when using just two (or finitely many) collapsing functions, but becomes important for infinitely many of them, is to define
\psi(\alpha)
0
1
\omega
\Omega
\Omega2
\psi1
\psi
\alpha
\psi1
\alpha
\psi(\Omega2)
\psi(\psi1(\Omega2))
\psi(\psi1(\Omega2))=\psi(\zeta\Omega+1)
\Omega2
\psi1
\Omega2
\psi
\Omega2
\psi1
\psi
\psi1
Indeed, there is no reason to stop at two levels: using
\omega+1
\Omega1,\Omega2,\ldots,\Omega\omega
\omega+1
1
\omega
\psi
\theta
\psi0(\varepsilon
\Omega\omega+1 |
)
1 | |
\Pi | |
1 |
Most definitions of ordinal collapsing functions found in the recent literature differ from the ones we have given in one technical but important way which makes them technically more convenient although intuitively less transparent. We now explain this.
The following definition (by induction on
\alpha
\psi
Let
C(\alpha,\beta)
0
1
\omega
\Omega
\beta
\psi\upharpoonright\alpha
\psi(\alpha)
\rho
C(\alpha,\rho)\cap\Omega=\rho
(This is equivalent, because if
\sigma
C(\alpha,0)
\psi(\alpha)
C(\alpha,0)=C(\alpha,\sigma)
\psi
\sigma
\Omega
C(\alpha,\sigma)
We can now make a change to the definition which makes it subtly different:
Let
\tildeC(\alpha,\beta)
0
1
\omega
\Omega
\beta
\tilde\psi\upharpoonright\alpha
\tilde\psi(\alpha)
\rho
\tildeC(\alpha,\rho)\cap\Omega=\rho
\alpha\in\tildeC(\alpha,\rho)
The first values of
\tilde\psi
\psi
\alpha<\zeta0
\zeta0=\varphi2(0)
\tilde\psi(\alpha)=\psi(\alpha)
\alpha\in\tildeC(\alpha,\rho)
\psi
\zeta0
\zeta0\leq\alpha\leq\Omega
\tilde\psi
\tilde\psi(\zeta0)=
\varepsilon | |
\zeta0+1 |
\alpha\in\tildeC(\alpha,\rho)
\tilde\psi(\zeta0)>\zeta0
\tilde\psi(\Omega)=\zeta0
\Omega\inC(\alpha,\rho)
\rho
\tilde\psi
\psi
Despite these changes, the
\tilde\psi
\psi(\Omega+1+\alpha)=\tilde\psi(\tilde\psi(\Omega)+\alpha)
\alpha
\psi(\Omega2)=\tilde\psi(\Omega+1)
Arai's ψ function is an ordinal collapsing function introduced by Toshiyasu Arai (husband of Noriko H. Arai) in his paper: A simplified ordinal analysis of first-order reflection.
\psi\Omega(\alpha)
\psi\Omega(\alpha)<\Omega
\Omega
KP\PiN |
\PiN |
KN
1 | |
\Pi | |
N-2 |
\PiN
N
\ge3
\Omega0=0
Suppose
KP\PiN |
\vdash\theta
\Sigma1 |
\Omega
\theta
n
\alpha=\psi\Omega(\omegan(KN+1))
L\alpha\models\theta
KP\PiN |
\{\alpha\inOT:\alpha<\psi\Omega(\omegan(KN+1))\};n=1,2,\ldots
\psi\Omega(\varepsilon
KN+1 |
)
KP\PiN |
\psi\Omega(\varepsilon\Omega)=|KP\omega|=BHO
\Omega
KP\omega
BHO
\psi\Omega(\Omega\omega)=
|
|
=BO
\Omega\omega
BO
\psi\Omega
(\varepsilon | |
\Omega\omega+1 |
)=|KPl|=TFBO
\Omega\omega
KPl
TFBO
\psi\Omega(\varepsilonI)=|KPi|
I
KPi
The first true OCF, Bachmann's
\psi
\Omega
\omega1
C\Omega(\alpha,\beta)
\beta\cup\{0,\Omega\}
(\xi → \omega\xi)
(\xi → \psi\Omega(\xi))
\xi<\alpha
\psi\Omega(\alpha)
C\Omega(\alpha,\rho)\cap\Omega=\rho
\psi\Omega(\varepsilon\Omega+1)
See main article: Buchholz psi functions. Buchholz's
\psi
\psi\nu:On → On
\psi\nu(\alpha)
\psi\nu\alpha
\Omega0=1
\Omega\nu=\aleph\nu
\nu>0
P(\alpha)
\alpha
\omega\xi
\xi\inOn
0 | |
C | |
\nu(\alpha) |
=\Omega\nu
n+1 | |
C | |
\nu(\alpha) |
=
n | |
C | |
\nu(\alpha) |
\cup\{\gamma\midP(\gamma)\subseteq
n | |
C | |
\nu |
(\alpha)\}\cup\{\psi\nu(\xi)\mid\xi\in\alpha\cap
n | |
C | |
\nu(\alpha) |
\land\xi\inCu(\xi)\landu\leq\omega\}
C\nu(\alpha)=cup\limitsn
n | |
C | |
\nu(\alpha) |
\psi\nu(\alpha)=min(\{\gamma\mid\gamma\notinC\nu(\alpha)\})
The limit of this system is
\psi0(\varepsilon
\Omega\omega+1 |
)
This OCF is a sophisticated extension of Buchholz's
\psi
\psi0(\Omega
|
)
\Omega | |||||
|
\Omega0=1
\Omega\nu=\aleph\nu
\nu>0
0 | |
C | |
\nu(\alpha) |
=\{\beta\mid\beta<\Omega\nu\}
n+1 | |
C | |
\nu(\alpha) |
=\{\beta+\gamma,\psi\mu(η)\mid\mu,\beta,\gamma,η\in
n | |
C | |
\nu(\alpha) |
\landη<\alpha\}
C\nu(\alpha)=cup\limitsn
n | |
C | |
\nu(\alpha) |
\psi\nu(\alpha)=min(\{\gamma\mid\gamma\notinC\nu(\alpha)\})
This OCF was the same as the ψ function previously used throughout this article; it is a simpler, more efficient version of Buchholz's ψ function defined by David Madore. Its use in this article lead to widespread use of the function.
C0(\alpha)=\{0,1,\omega,\Omega\}
Cn+1(\alpha)=\{\gamma+\delta,\gamma\delta,\gamma\delta,\psi(η)\mid\gamma,\delta,η\inCn(\alpha);η<\alpha\}
C(\alpha)=cup\limitsnCn(\alpha)
\psi(\alpha)=min(\{\beta\in\Omega\mid\beta\notinC(\alpha)\})
This function was used by Chris Bird, who also invented the next OCF.
Chris Bird devised the following shorthand for the extended Veblen function
\varphi
\theta(\Omegan-1an-1+ … +
2a | |
\Omega | |
2 |
+\Omegaa1+a0,b)=\varphi(an-1,\ldots,a2,a1,a0,b)
\theta(\alpha,0)
\theta(\alpha)
This function is only defined for arguments less than
\Omega\omega
Jäger's ψ is a hierarchy of single-argument ordinal functions ψκ indexed by uncountable regular cardinals κ smaller than the least weakly Mahlo cardinal M0 introduced by German mathematician Gerhard Jäger in 1984. It was developed on the base of Buchholz's approach.
\kappa=I\alpha(0)
\kappa-=0
\kappa=I\alpha(\beta+1)
\kappa-=I\alpha(\beta)
0 | |
C | |
\kappa(\alpha) |
=\{\kappa-\}\cup\kappa-
n+1 | |
C | |
\kappa(\alpha) |
\subsetM0
n | |
C | |
\kappa(\alpha) |
\subsetM0
n+1 | |
C | |
\kappa(\alpha) |
\subsetM0
\beta,\gamma\in
n | |
C | |
\kappa(\alpha) |
\varphi\beta(\gamma)\in
n+1 | |
C | |
\kappa(\alpha) |
\beta,\gamma\in
n | |
C | |
\kappa(\alpha) |
I\beta(\gamma)\in
n+1 | |
C | |
\kappa(\alpha) |
\pi\in
n | |
C | |
\kappa(\alpha) |
\gamma<\pi<\kappa ⇒ \gamma\in
n+1 | |
C | |
\kappa(\alpha) |
\gamma\in\alpha\cap
n | |
C | |
\kappa(\alpha) |
\pi\in
n | |
C | |
\kappa(\alpha) |
\gamma\inC\pi(\gamma) ⇒ \psi\pi(\gamma)\in
n+1 | |
C | |
\kappa(\alpha) |
C\kappa(\alpha)=cup\limitsn
n | |
C | |
\kappa(\alpha) |
\psi\kappa(\alpha)=min(\{\xi\in\kappa\mid\xi\notinC\kappa(\alpha)\})
This is a sophisticated simplification of Jäger's ψ created by Denis Maksudov. An ordinal is α-weakly inaccessible if it is uncountable, regular and it is a limit of γ-weakly inaccessible cardinals for γ < α. Let I(α, 0) be the first α-weakly inaccessible cardinal, I(α, β + 1) be the first α-weakly inaccessible cardinal after I(α, β) and I(α, β) =
sup(\{I(\alpha,\gamma)\mid\gamma<\beta\})
C0(\alpha,\beta)=\beta\cup\{0\}
Cn(\alpha,\beta)=\{\gamma+\delta\mid\gamma,\delta\inCn(\alpha,\beta)\}\cup\{I(\gamma,\delta)\mid\gamma,\delta\inCn(\alpha,\beta)\}\cup\{\psi\pi(\gamma)\mid\pi,\gamma,\inCn(\alpha,\beta)\land\gamma<\alpha\}
C(\alpha,\beta)=cup\limitsnCn(\alpha,\beta)
\psi\pi(\alpha)=min(\{\beta<\pi\midC(\alpha,\beta)\cap\pi\subseteq\beta\})
See main article: Rathjen's psi function. Rathjen's Ψ function is based on the least weakly compact cardinal to create large countable ordinals. For a weakly compact cardinal K, the functions
M\alpha
C(\alpha,\pi)
\Xi(\alpha)
\xi | |
\Psi | |
\pi(\alpha) |
K\capLim
\{\pi<K\midC(\alpha,\pi)\capK=\pi\land\forall\xi\inC(\alpha,\pi)\cap\alpha,M\xi
\pi\land\alpha\inC(\alpha,\pi)\}
C(\alpha,\beta)
\beta\cup\{0,K\}
(\xi,η) → \varphi(\xi,η)
\xi → \Omega\xi
\xi → \Xi(\xi)
(\xi,\pi,\delta) →
\xi | |
\Psi | |
\pi(\delta) |
\xi\leq\delta<\alpha
\Xi(\alpha)=min(M\alpha\cup\{K\})
\xi\leq\alpha
\xi | |
\Psi | |
\pi(\alpha) |
=min(\{\rho\inM\xi\cap\pi:C(\alpha,\rho)\cap\pi=\rho\land\pi,\alpha\inC(\alpha,\rho)\}\cup\{\pi\})
As noted in the introduction, the use and definition of ordinal collapsing functions is strongly connected with the theory of ordinal analysis, so the collapse of this or that large cardinal must be mentioned simultaneously with the theory for which it provides a proof-theoretic analysis.
1 | |
\Delta | |
2 |
\alpha\mapsto\Omega\alpha
C( ⋅ )
\Pi3
\Xi(\alpha)
\alpha
\alpha\mapsto\Xi(\alpha)
\vec\xi | |
\psi | |
\pi |
\xi
1 | |
\Pi | |
n |
n>0
\Pin+2
1 | |
\Pi | |
2 |
\Sigma1
1 | |
\Delta | |
2 |
1 | |
\Pi | |
2 |