Arrow Relation in Extremal Set Theory

• Arrow relation is a quantitative statement in extremal set theory that asserts any family of subsets above a certain size must exhibit a rich trace on some smaller set.
• It employs hereditary reduction and the squash operation to transform any set family into a hereditary one without increasing local trace complexity.
• The concept bridges key ideas such as Sauer–Shelah shattering, VC-dimension, and hypergraph Turán problems, offering exact thresholds and asymptotic insights.

In extremal set theory, the arrow relation is a quantitative statement about how a large family of subsets of $[n]$ must induce a rich system of traces on some smaller ground set. In the standard notation,

$(n,m)\rightarrow(a,b),$

the meaning is that for every family $\mathcal F\subseteq 2^{[n]}$ with $|\mathcal F|\ge m$, there exists a subset $T\subseteq [n]$ with $|T|=a$ such that the trace

$\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$

has size at least $b$. The concept generalizes Sauer–Shelah type shattering statements and connects trace theory, VC-dimension, forbidden configurations, and hypergraph extremal problems (Li et al., 31 Jul 2025).

1. Definition and basic interpretation

Let $[n]=\{1,2,\dots,n\}$, and let $\mathcal F\subseteq 2^{[n]}$. For a set $(n,m)\rightarrow(a,b),$0, the trace of $(n,m)\rightarrow(a,b),$1 on $(n,m)\rightarrow(a,b),$2 is

$(n,m)\rightarrow(a,b),$3

Thus the trace records all intersection patterns realized by members of $(n,m)\rightarrow(a,b),$4 on the coordinates in $(n,m)\rightarrow(a,b),$5.

The arrow notation

$(n,m)\rightarrow(a,b),$6

means that every family $(n,m)\rightarrow(a,b),$7 with $(n,m)\rightarrow(a,b),$8 has an $(n,m)\rightarrow(a,b),$9-element subset $\mathcal F\subseteq 2^{[n]}$0 on which at least $\mathcal F\subseteq 2^{[n]}$1 distinct traces occur. The survey also uses the shorthand $\mathcal F\subseteq 2^{[n]}$2 when there exists such a set $\mathcal F\subseteq 2^{[n]}$3, and $\mathcal F\subseteq 2^{[n]}$4 when every $\mathcal F\subseteq 2^{[n]}$5-subset $\mathcal F\subseteq 2^{[n]}$6 satisfies $\mathcal F\subseteq 2^{[n]}$7.

If $\mathcal F\subseteq 2^{[n]}$8, then $\mathcal F\subseteq 2^{[n]}$9, so at most $|\mathcal F|\ge m$0 patterns can occur. The extreme case $|\mathcal F|\ge m$1 is the usual notion of shattering: every subset of $|\mathcal F|\ge m$2 appears as $|\mathcal F|\ge m$3 for some $|\mathcal F|\ge m$4. Arrow relations therefore interpolate between weak local richness and full shattering.

2. Foundational results and hereditary reduction

The foundational example is the Sauer–Shelah lemma. In arrow notation it states that

$|\mathcal F|\ge m$5

equivalently,

$|\mathcal F|\ge m$6

This is the threshold forcing some $|\mathcal F|\ge m$7-set to be shattered.

A central reduction due to Frankl is that arrow relations may be tested on hereditary families. A family $|\mathcal F|\ge m$8 is hereditary if $|\mathcal F|\ge m$9 implies $T\subseteq [n]$0. Frankl’s lemma states that the following are equivalent:

1. $T\subseteq [n]$1.
2. Every hereditary family $T\subseteq [n]$2 with $T\subseteq [n]$3 satisfies $T\subseteq [n]$4.

The reduction is implemented by the squash operation $T\subseteq [n]$5, which replaces certain sets $T\subseteq [n]$6 by $T\subseteq [n]$7 when the latter is absent. Repeated squashing transforms an arbitrary family into a hereditary one without increasing traces, because

$T\subseteq [n]$8

This is one of the main structural tools in the subject: extremal obstructions to an arrow relation can be taken hereditary.

3. Defect Sauer problems and exact thresholds

For fixed $T\subseteq [n]$9, the principal threshold parameter is

$|T|=a$0

The case $|T|=a$1 is the Sauer–Shelah lemma. The regime $|T|=a$2 is often called the defect Sauer problem, and many exact values are known (Li et al., 31 Jul 2025).

Parameter	Exact value
$|T|=a$3	$|T|=a$4
$|T|=a$5	$|T|=a$6
$|T|=a$7	$|T|=a$8
$|T|=a$9, $\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$0	$\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$1
$\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$2, $\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$3	$\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$4

The case

$\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$5

is the prototype of the method. After hereditary reduction, a family $\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$6 cannot contain a $\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$7-set, because any $\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$8-set in a hereditary family yields full trace $\mathcal F_{|T}=\{F\cap T:F\in\mathcal F\}$9 on itself. Hence all members have size at most $b$0, and the $b$1-sets form a triangle-free graph. Mantel’s theorem then bounds their number by $b$2. The lower bound is witnessed by

$b$3

where $b$4 is the balanced complete bipartite graph.

Several four-vertex cases remain open. For $b$5, the problem reduces to $b$6, and the survey gives

$b$7

For $b$8 and $b$9, the values depend on $[n]=\{1,2,\dots,n\}$0 and $[n]=\{1,2,\dots,n\}$1, with bounds

$[n]=\{1,2,\dots,n\}$2

$[n]=\{1,2,\dots,n\}$3

These cases show that defect Sauer questions quickly meet difficult hypergraph Turán problems.

4. Single-element removal and the asymptotic slope $[n]=\{1,2,\dots,n\}$4

A second major direction studies the relations

$[n]=\{1,2,\dots,n\}$5

Here $[n]=\{1,2,\dots,n\}$6 denotes the maximum $[n]=\{1,2,\dots,n\}$7 for which this implication holds. By hereditary reduction, this is equivalent to the statement that every hereditary family $[n]=\{1,2,\dots,n\}$8 on $[n]=\{1,2,\dots,n\}$9 vertices with $\mathcal F\subseteq 2^{[n]}$0 edges has minimum degree $\mathcal F\subseteq 2^{[n]}$1.

Two exact finite-$\mathcal F\subseteq 2^{[n]}$2 results are classical: $\mathcal F\subseteq 2^{[n]}$3

Frankl proved that for positive integers $\mathcal F\subseteq 2^{[n]}$4,

$\mathcal F\subseteq 2^{[n]}$5

When $\mathcal F\subseteq 2^{[n]}$6, a block construction gives equality: $\mathcal F\subseteq 2^{[n]}$7

Watanabe and Frankl proved that the limit

$\mathcal F\subseteq 2^{[n]}$8

exists. They obtained

$\mathcal F\subseteq 2^{[n]}$9

Recent work determines the asymptotic slope near powers of two with high precision. The survey records:

• Piga and Schülke (2021):

$(n,m)\rightarrow(a,b),$00

• Li, Ma, Rong (2024):

$(n,m)\rightarrow(a,b),$01

and for $(n,m)\rightarrow(a,b),$02, $(n,m)\rightarrow(a,b),$03,

$(n,m)\rightarrow(a,b),$04

• Reiher and Schülke (2025):

$(n,m)\rightarrow(a,b),$05

The survey also lists exact values for $(n,m)\rightarrow(a,b),$06, for example

$(n,m)\rightarrow(a,b),$07

The current picture is therefore sharp just below powers of two, but substantially less complete away from that regime.

5. Trace functions, VC-dimension, and extremal families

Instead of fixing $(n,m)\rightarrow(a,b),$08 and asking for the minimum $(n,m)\rightarrow(a,b),$09, one may fix $(n,m)\rightarrow(a,b),$10 and ask for the largest $(n,m)\rightarrow(a,b),$11 guaranteed. For a family $(n,m)\rightarrow(a,b),$12, define

$(n,m)\rightarrow(a,b),$13

and

$(n,m)\rightarrow(a,b),$14

This is exactly the largest $(n,m)\rightarrow(a,b),$15 such that $(n,m)\rightarrow(a,b),$16 holds (Li et al., 31 Jul 2025).

A universal lower bound due to Bollobás and Radcliffe is

$(n,m)\rightarrow(a,b),$17

They also proved sharper bounds for polynomial-size families and linear-size traces. For integer $(n,m)\rightarrow(a,b),$18 and $(n,m)\rightarrow(a,b),$19,

$(n,m)\rightarrow(a,b),$20

where $(n,m)\rightarrow(a,b),$21, and

$(n,m)\rightarrow(a,b),$22

Kahn, Kalai and Linial proved that for $(n,m)\rightarrow(a,b),$23,

$(n,m)\rightarrow(a,b),$24

where $(n,m)\rightarrow(a,b),$25.

The survey also emphasizes that initial segments of the Boolean lattice do not minimize traces in general. Bollobás and Radcliffe constructed, for any $(n,m)\rightarrow(a,b),$26 and $(n,m)\rightarrow(a,b),$27, a hereditary family $(n,m)\rightarrow(a,b),$28 with

$(n,m)\rightarrow(a,b),$29

such that

$(n,m)\rightarrow(a,b),$30

Hence

$(n,m)\rightarrow(a,b),$31

A major recent advance due to Alon, Moshkovitz, and Solomon states that if $(n,m)\rightarrow(a,b),$32, then

$(n,m)\rightarrow(a,b),$33

where

$(n,m)\rightarrow(a,b),$34

Moreover, if $(n,m)\rightarrow(a,b),$35 and $(n,m)\rightarrow(a,b),$36, then

$(n,m)\rightarrow(a,b),$37

For $(n,m)\rightarrow(a,b),$38 and $(n,m)\rightarrow(a,b),$39,

$(n,m)\rightarrow(a,b),$40

On the structural side, traced and strongly traced sets lead to extremal equalities. Let

$(n,m)\rightarrow(a,b),$41

and define the VC-dimension by

$(n,m)\rightarrow(a,b),$42

Pajor proved

$(n,m)\rightarrow(a,b),$43

A set $(n,m)\rightarrow(a,b),$44 is strongly traced by $(n,m)\rightarrow(a,b),$45 if there exists $(n,m)\rightarrow(a,b),$46 such that

$(n,m)\rightarrow(a,b),$47

If $(n,m)\rightarrow(a,b),$48 denotes the family of strongly traced sets, then Bollobás, Leader, and Radcliffe proved the reverse Sauer inequality

$(n,m)\rightarrow(a,b),$49

A family is s-extremal if

$(n,m)\rightarrow(a,b),$50

Bollobás and Radcliffe showed that this is equivalent to several other properties, including

$(n,m)\rightarrow(a,b),$51

and the uniqueness of the hereditary family obtained from $(n,m)\rightarrow(a,b),$52 by repeated squashing. A related strengthening, order-shattering, satisfies the exact identity

$(n,m)\rightarrow(a,b),$53

6. Matrix and hypergraph reformulations, and open directions

Arrow relations admit natural reformulations in the language of forbidden $(n,m)\rightarrow(a,b),$54-$(n,m)\rightarrow(a,b),$55 matrices and hypergraph traces (Li et al., 31 Jul 2025). Every family $(n,m)\rightarrow(a,b),$56 corresponds to a simple $(n,m)\rightarrow(a,b),$57-$(n,m)\rightarrow(a,b),$58 incidence matrix; traces then become configurations in submatrices. For a fixed $(n,m)\rightarrow(a,b),$59-$(n,m)\rightarrow(a,b),$60 matrix $(n,m)\rightarrow(a,b),$61, let $(n,m)\rightarrow(a,b),$62 be the maximum number of columns in an $(n,m)\rightarrow(a,b),$63-rowed simple matrix avoiding $(n,m)\rightarrow(a,b),$64 as a configuration.

In this framework, the complete $(n,m)\rightarrow(a,b),$65-row configuration $(n,m)\rightarrow(a,b),$66 yields Sauer–Shelah in the form

$(n,m)\rightarrow(a,b),$67

Related exact results include

$(n,m)\rightarrow(a,b),$68

$(n,m)\rightarrow(a,b),$69

and for general $(n,m)\rightarrow(a,b),$70,

$(n,m)\rightarrow(a,b),$71

The hypergraph formulation uses induced Berge copies. For a graph $(n,m)\rightarrow(a,b),$72, let $(n,m)\rightarrow(a,b),$73 denote the family of $(n,m)\rightarrow(a,b),$74-uniform induced Berge copies of $(n,m)\rightarrow(a,b),$75. Füredi and Luo proved that for every graph $(n,m)\rightarrow(a,b),$76 with at least one edge,

$(n,m)\rightarrow(a,b),$77

Mubayi and Zhao determined asymptotics for several clique cases: if $(n,m)\rightarrow(a,b),$78 or $(n,m)\rightarrow(a,b),$79, then

$(n,m)\rightarrow(a,b),$80

The main open problems surveyed in this area are structural as well as enumerative. They include the Frankl–Wang conjecture

$(n,m)\rightarrow(a,b),$81

Frankl’s conjecture on shattering in antichains, the Mészáros–Rónyai conjecture that every nonempty s-extremal family contains an element whose removal preserves s-extremality, the exact values of $(n,m)\rightarrow(a,b),$82 and $(n,m)\rightarrow(a,b),$83, and the determination of $(n,m)\rightarrow(a,b),$84 away from powers of two.

In this form, the arrow relation functions as a unifying language for extremal trace phenomena. It quantifies how global size in $(n,m)\rightarrow(a,b),$85 forces local complexity, and it does so in a way that naturally interfaces with hereditary set systems, VC-theory, Turán-type bounds, forbidden matrices, and induced hypergraph traces.