PROFESIONAL EN LA ASIGNATURA “LA FÍSICA DEL PREUNIVERSITARIO Y SU METODOLOGÍA”
CAPÍTULO 2: UNA METODOLOGÍA PARA UTILIZAR LAS TAREAS DOCENTES CON ENFOQUE SOCIOCULTURAL-PROFESIONAL EN LA ASIGNATURA “LA
4. La didáctica general como modelo teórico generalizador para el desarrollo de la asignatura “La Física del Preuniversitario y su Metodología”
2.2.3 Etapas para utilizar las tareas docentes con enfoque sociocultural-profesional
π−1((
−∞,−3]∪[3,∞))⊂P(),
whereπ:C(S)→Zis the projection to the annulus with core curve v(∗κ)post-composed with a transla- tion such thatπ(b)=0.
Proof. Letκbe any sequence starting and ending atφ0such that every entry in the associated matrixQκ is at least 3. This implies that the stable vertex cyclev(∗κ) passes over every band inat least thrice.
Now consider the subsurface projectionπto the annulus with core curvev(∗κ). By post-composing with a translation ofZ, we may assume thatπ(b)=0. By Proposition 12.2, the projection πof every point inC(S) \P() is within distance 3 ofπ(b). This implies that the pre-imageπ−1((−∞,−3)∪(3,∞)) is
In other words, when the vertex cyclev(∗κ) gets nested deep enough inP() a fixed pre-image of the projection tov(∗κ) is contained inP(). It is in this sense that such aκis apush-in sequence. We fix such a sequenceκonce and for all. This achieves a setup to which Theorem 11.6 can be applied.
13. ASINGULAR SET FOR HARMONIC MEASURE.
13.1. Choosing the initial chart. We choose the initial embedding ofφ0inSwith the underlying trackτ0 such that the base-pointb∈C(S) does not belong toP(τ0). The embedding identifies the configuration spaceW0with a chart inP M F. The construction of a singular set will proceed inside this chart.
13.2. Relative probability that the sequenceκ∗n0follows aC-distributed stage. By Lemma 6.7, the relative probability that the push-in sequenceκfollows aC-distributed stage is up to a uniform con- stantc>1, the same as the probability`(PQκ(W0)) that an expansion begins withκi.e.,
1
c`(PQκ(W0))<
`(PQ∗κ(W0)) `(PQ(W0)) <
c·`(PQκ(W0)).
Sinceκis a priori fixed, whenever the sequenceκ follows aC-distributed stage , the resulting stage
∗κisC0-distributed, for someC0that depends only onCandd. Recall from Section 7 that0denotes the Dehn twist sequence. Denote the matrix associated to the sequence ∗κ∗n0byQ,n. Again by Lemma 6.7, there exists a constanta1that depends only ona0andC0, such that
1 a1nj < `(PQ,n(W0)) `(PQ∗κ(W0))< a1 nj.
Hence the relative probability that the sequenceκ∗n0follows aC-distributed stagesatisfies
(13.3) 1 c a1nj <`(PQ,n(W0)) `(PQ(W0)) < c a1 nj i.e., the relative probability`(PQ,n(W0))/`(PQ(W0))≈1/nj.
13.4. Construction of the doubly indexed sequence of setsYn(m). By Theorem 6.9, almost every expan- sion becomesC-distributed infinitely often. Hence, for every non-negative integerm, almost every ex- pansion has a stage that is them-th instance ofC-distribution. Whenm=0, we mean the initial stage itself with no splitting whatsoever.
LetSm be the set of stages that are them-th instances ofC-distribution. For ∈Sm, letY be the set ofxinW0whose expansion begins with i.e., the subsetY=PQ(W0). For distinct , ¯ inSm, the setsY,Y¯have disjoint interiors. By Theorem 6.9, the union over all∈Sm of the setsYis a set of full measure i.e.,
(13.5) X
∈Sm
`(Y)=1
Follow each inSm by the sequence κ∗n0, and letQ,n be the matrix associated to ∗κ∗n0. Let Y,n=PQ,n(W0). Then, the ratio`(Y,n)/`(Y) satisfies Estimate (13.3). LetYn(m)be the union
Yn(m)= [
∈Sm Y,n.
Form=0, the setYn(0)is justPQn0(W0) i.e., the set ofxwhose expansion begins withn0. First, we
estimate the Lebesgue measure ofYn(m).
Lemma 13.6. (13.7) 1 c a1nj <`(Yn(m))< c a1 nj
Proof. Write `(Yn(m))= X ∈Sm `(Y,n)= X ∈Sm `(Y)`(Y,n) `(Y) .
Estimate (13.3) for the ratio`(Y,n)/`(Y), and Equation (13.5) finishes the proof. In particular, the above lemma shows that for anym1,m2,
(13.8) `(Y(m1)
n )≈`(Yn(m2))
13.9. Almost independence ofYn(m)for the Lebesgue measure. Let (m1,n1) and (m2,n2) be a pair of
indices withm1<m2. Since the setsYι:ι∈Sm1is a partition of a set of full measure, it is enough to check
that almost independence holds in eachYι. For anyι∈Sm1 and∈Sm2, eitherYis contained inYι,n1or
has interior disjoint from it. By Lemma 6.7, givenι∈Sm1, the relative probabilities satisfy
Prob(Y(m1) n1 |ι) ≈ `(Y (0) n1 ) Prob(Y(m2) n2 |ι) ≈ `(Y (m2−m1) n2 ) Prob(Y(m1) n1 ∩Y (m2) n2 |ι) ≈ `(Y (0) n1 ∩Y (m2−m1) n2 )
So it is enough to check that for anym>0, the setsYn(0)1 andY
(m)
n2 are almost independent. Again, the
main point is that for allinSm, the setY is either contained inYn(0)1 or has interior disjoint fromY
(0)
n1 .
LetTm be the subset ofSm consisting of those for whichY is contained inYn(0)1 . The union over all
∈Tmof the setsYis a set of full measure inYn(0)1 . Hence, using Estimate (13.3) and Equation (13.8), we
get `¡ Yn(0)1 ∩Y (m) n2 ¢ ≈ X ∈Tm `(Y,n2) ≈ X ∈Tm `(Yn(0)2 )`(Y) ≈ `(Yn(m2 )) X ∈Jm `(Y) ≈ `(Yn(m2 ))`(Yn(0)1 ) showing almost independence.
13.10. Harmonic measure estimate forYn(m). Recall from Remark 7.3 thatφ0has the property that sim- ple closed curves carried by the underlying train track twist only to the left around the stable vertex cycle v. Letπbe the projection to the annulus with core curvev. Since the twisting is in one direction only, the projection of the set of carried curves is a one-sided interval of the form [M1,∞) or (−∞,−M1]. Reversing
the order onZif necessary, we fix the convention that it is of the form [M1,∞).
For the initial trackτ0, letπbe the projection to the annulus with core curvev(κ). Post-composing by a translation ofZassume thatπ(b)=0. By our chosen convention, there is a positive integerM1>3
such thatπ(P(κ))=[M1,∞).
For∈Sm, letv(∗κ) be the stable vertex cycle of ∗κ. Denote byπ∗κ, the sub-surface projection to the annulus with core curvev(∗κ). Post-composing by a translation, we assume thatπ∗κ(b)=0.
By Lemma 12.6, the push-in sequenceκensures that the pre-imageπ−1
∗κ([3,∞)∪(−∞,−3]) sits entirely inside the setP(). By our chosen convention,π∗κ(P(∗κ)) is a one sided interval. We show below that this interval is contained in [M1−3,∞)