1. Introduction
Let be an integral and non-degenerate n-dimensional variety.
To recall the classical notion of abstract secant variety, its map to and its differential (computed in geometric term by A. Terracini) we use the following notation.
For any closed subscheme let denote its linear span. Let denote the set of all smooth points of X. For any let (or just ) denote the closed subscheme of X with as its ideal sheaf. The scheme is a zero-dimensional scheme and . Moreover, is the Zariski tangent space of X at o. For all finite subsets set .
Fix a positive integer
. Let
denote the set of all
such that
and
S is linearly independent. The set
is a smooth quasi-projective variety and its closure
in the Hilbert scheme of
X is an integral projective variety, singular if
. Let
denote the set of all pairs
such that
and
for all
. The set
is a smooth quasi-projective variety. The closure
of
in
is a closed and irreducible projective variety, often called the abstract
s-secant variety of
X. Call
the morphism induced by the projection
. The irreducible variety
is the
s-secant variety of
X, i.e., the closure in
of the union of all linear spaces
for some subset of
X with cardinality
s. However, since
is usually very singular (even for nice
X) we consider the differential of
only at the points of
. Set
. Fix
. A. Terracini proved that the image of the differential
of
at
does not depend on
q: it is the linear span of all tangent spaces
,
, i.e., it is
([
1], Cor. 1.11). Since Alessandro Terracini’s classical papers the study of the differential of
gave (for very good reasons: the results were both nice and useful) several papers, most of them for the case
S general in
X so that the rank of
is the integer
. Some recent papers also considered the case in which
S is not general ([
2,
3]). Here we consider both cases,
S general and
S very specific. An
arrow of
is a connected degree 2 scheme
. Since
v is assumed to be connected, its reduction
is a point,
. The Zariski tangent space
of
X is the union of all arrows of
X. Thus, the linear span
has the expected dimension if and only if all unions of
s arrows, each of them with as its support a different point of
S, are linearly independent. This observation due to K. Chandler ([
4,
5,
6]) had many applications ([
6,
7,
8]). A union of points and arrows may also be used to describe tensors and homogeneous polynomials. For instance in the additive decomposition of degree
d forms an arrow corresponds to a term in the sum of the form
, where
ℓ and
are non-proportional linear forms, while a point corresponds to an addendum
,
.
We consider a general
S for the Veronese embeddings
of
, but we ask if
is scheme-theoretically the scheme
or not. This is tricky and we discuss in more details why it is tricky in
Section 4. For all positive integers
n and
d let
(resp.
) be the maximal integer
such that
and
is globally generated (resp.
and
has no base points outside
), where
Z is a general union of
x double points of
. These integers may be expressed with the geometric language used for the additive decomposition of degree
d forms in the following way. Let
,
, denote the order
d Veronese embedding of
, i.e., the embedding given evaluating all degree
d forms in
variables. Set
. The integer
is the maximal integer
such that
and
scheme-theoretically.
We prove the following result.
Theorem 1. Fix integers and z such that . Let be a general union of z double points. Then and is globally generated.
The vanishing of
is well-known, but it is put in the statement because for a non-general
E (call if
F) it is easy to obtain
globally generated for some
F such that
. To prove Theorem 1 we use a degeneration of several planar double points ([
9]).
A natural question is the extension of Theorem 2 to the case . We ask also for the proof of similar results for tensors of certain formats and certain tensor ranks and for the case of partially symmetric tensors.
In
Section 5 we consider the following uniqueness problem for tensors. For which formats there is a concise tensor which can be irredundantly determined by the union of a point and an arrow for more than one union of a point and an arrow? Propositions 1 and 2 and Theorem 3 give the list of the exceptional cases.
The last section is speculative. Suppose that at a certain the differential of the Terracini map has a kernel. Is there a condition (using higher derivatives) which says if the fiber of at is positive-dimensional?
Unless otherwise stated we work over an algebraically closed field of characteristic 0.
3. Veronese Varieties
In this section, we prove Theorem 2. In
Section 4 we discuss why it does not follow from known results on weak nondefectivity and tangenential nondefectivity ([
10,
11,
12,
13]).
Obviously for all .
For all positive integers
n and
d let
denote the maximal cardinality of a finite set
such that
. Obviously
and
. Since the singular locus of a quadric hypersurface is a linear space,
. A key theorem due to Alexander and Hirschowitz ([
14,
15,
16,
17]) computes
for all
and says that
for all
, unless
. Moreover,
,
,
and
.
Remark 2. Obviously . Please note that if (and hence is not the trivial line bundle) and for all and , except in the 4 exceptional cases .
Remark 3. Let be a general subset such that . The Castelnuovo–Mumford lemma gives that is spanned and that . Thus, .
Remark 4. Since any 2 points of a projective space are collinear, . Take and . Since 4 general points of are the complete intersection of 2 conics and , we have . Thus, in a few cases the upper bound in Remark 3 is sharp.
Lemma 1. .
Proof. Since and , . Thus, it is sufficient to prove that . Fix a smooth cubic and take such that . Take such that o is not contained in any line spanned by 2 of the points of . We need to prove that has no base points outside S and that . We have . Since is very ample, the residual exact sequence of D shows that has no base locus outside . Thus, it is sufficient to use that is projectively normal and that . □
To prove Theorem 1 we prove the following result.
Theorem 2. Fix integers and z such that . Then there is a connected zero-dimensional scheme such that Z is a flat limit of a family of z pairwise disjoint double points, and has no base points outside .
Lemma 2. Let be a smooth curve. Fix and positive integers such that . Then there is a zero-dimensional scheme such that , , Z is a flat limit of a family of unions of z pairwise disjoint double points, and .
Proof. If
we use ([
18], Proposition 5.1.2), which corresponds in the set-up of [
9] to the front collision, which just adds the escaliers of the
z double points. Assume
and set
. Please note that
. We take a specialization
Z of a family of general unions
with
B a general unions of
double points of
with as reduction general points of
C and
E a general union of
e double points of
. Then we apply
e times (each time to a different connected component of
E) ([
18], §5.2) with the escalier
of a double point of
. □
Proof of Theorem 2: Let t be the maximal positive integer such that and .
Proof of Claim 1: Assume . Since t is maximal, , i.e., , i.e., . We have , i.e., , contradicting the assumption .
Fix a general
. Thus,
C is smooth and
. Set
. By Lemma 2 there is a connected zero-dimensional scheme
such that
and
. Thus,
is a general connected zero-dimensional subscheme of
C of degree
. By [
19]
and
. Since
and
C is projectively normal,
. The residual exact sequence of
C gives
and that the restriction map
is surjective. Call
W the scheme-theoretic base locus of
. Since
C is a smooth curve and
is a general connected degree
w subscheme of
C,
20] gives that no point of
is a base point of
. The surjectivity of
gives that no point of
is a base point of
. Since a general osculating space of a curve is not hyperosculating and
is surjective, we obtain
as schemes. Since
,
. Since
is a general connected degree
subscheme of the smooth curve
C and
,
20] gives
. Thus,
has no base point outside
C. Since no point of
is a base point of
, no
is a base point of
. □
Proof of Theorem 1: Being globally generated is an open condition in families of coherent sheaves with constant cohomology. Thus, it is sufficient to prove that if
the scheme
Z constructed in the proof of Theorem 2 is globally generated. Let
be the image of the evaluation map
. Since
, to conclude the proof it is sufficient to prove that
. Since
has no base points outside
,
W is a connected zero-dimensional scheme. Take
C,
t and
w as in the proof of Theorem 2. We saw that
. Since
,
. Please note that
is a general connected zero-dimensional scheme of degree
. Hence
has no base points ([
20]; note that in [
20] “curve” means “smooth curve”). Consider the residual exact sequence
We saw that and that and are globally generated. Thus, is globally generated. □
4. Base Point Freeness
In this section, we point out why it is still open, even for generic symmetric rank, although very similar statements are true, stated and proved in the literature ([
10,
11,
12,
13]).
Assume characteristic zero.
Let
be an integral and non-degenerate variety. Set
and fix an integer
such that
and
, i.e.,
X is not secant
s-defectivity. Fix a general
such that
and set
. Since
,
has dimension
. Take a general hyperplane
containing
Z. There are key notions due to C. Ciliberto and L. Chiantini (weak nondefectivity and tangential nondefectivity) ([
10,
11,
12]) which when they are satisfies imply that
H is tangent to
only at the points of
S. This true statement does not imply that
meets
only at
S and the linear spaces contained in
X and containing at least one point of
S, even when
has codimension
. It would seem intuitively true, but it is only conjectural for curves and false in some cases for higher-dimensional smooth manifolds ([
21]).
We say that Assumption 1 holds if the following conjectural statement is true:
Assumption 1. Assume . Let , , be any integral and non-degenerate curve. For a general the tangent line of X at p meets X only at p.
An assumption similar to Assumption 1 trivially fail for all
X such that
and
X is covered by lines. However, an example due to M. Ohno shows that it may fail even for smooth manifolds of general type [
21]. See [
22,
23] for many partial solutions and applications of Assumption 1. Thus, we cannot freely extend to general unions of double points the following observation concerning general finite sets.
Remark 5. Let be an integral and non-degenerate variety. Set . Let be a general set such that . In characteristic 0 an easy application of the linear general position of a general codimension n linear section gives (scheme-theoretic intersection).
For any
and all positive integer
m let
denote the closed subscheme of
with
as its ideal sheaf. The linear space
is the
m-osculating linear space of
X at
o. Motivated by [
20] we consider the following Condition (Assumption 2):
Assumption 2. Assume . Fix integers . Let be any integral and non-degenerate curve. For a general the m osculating space of X at p meets X only at p and the support of the union of the point and the arrow. Usually, this support is a very small part of the Segre variety.
We are working over an algebraically closed field of characteristic zero, because Assumption 1 fails in positive characteristic even for some smooth curves ([
24], Example 4.1). In characteristic 0 Assumptions 1 and 2 holds for all smooth curves ([
20,
23,
24], Theorem 3.1)
5. Tensors
Tensors associated with an arrow are exactly the tensors contained in the tangential variety of the Segre variety related to the format of the tensor. In this section we describe all concise tensors which are linear combinations in two different ways of a rank 1 tensor and a tensor associated with an arrow (Theorem 3). Theorem 3 lists 7 cases with for each case a quotation of an example or remark of the paper. The remark or example describes in detail each exceptional case. In each case we describe by how many parameters the possible unions of an arrow and a point depend.
We recall the following properties of the Segre varieties and their connection with tensors and the tensor rank of a tensor ([
25]).
Let , , , , be a multiprojective space. Let denote the projection of Y onto its i-th factor. If set and let denote the projection. The map is the map forgetting the i-th coordinate of each . Let denote the Segre embedding of .
For all set . The line bundles , , form a -basis of the abelian group . The Künneth formula gives if some , if for all i and if for all i. For any let (resp. ) be the line bundle on Y with multidegree with and for all (resp. We have . Set . Let denote the Segre embedding of Y.
For any tensor T of format , , the tensor rank of T is the -rank of the element associated with T.
The main result of this section is the following one.
Theorem 3. Let be a union of a point and an arrow. Assume that Y is the minimal multiprojective space containing Y. Fix such that for any . Assume the existence of a union of a point and an arrow such that and . Then Y is as in one of the following cases (assuming for all ):
- 1 .
and either , or and .
- 2 .
and either , or and .
- 3 .
, and are as in Remark 10.
- 4 .
, and are as in Remark 11.
- 5 .
, and are as in Example 9.
- 6 .
, and A, B are as in Example 3.
- 7 .
and is as in Examples 4, 5, 6, 7 or 8.
In each case we give a rough description of the possible B’s. The one with (resp. ) are listed in Remark 10 (resp. 11). The last case, i.e., Examples 4, 5, 7 or 8 are the only cases which allow any .
Remark 6. Let be a zero-dimensional scheme of degree . The minimal multiprojective subspace of Y containing Z is the multiprojective space . Write for some positive integers s and , . We have for all i and this is in general the only restriction we may obtain from the isomorphism class of Z as an abstract scheme. Of course, and if the h-th positive-dimensional factor of is contained in the i-th-dimensional factor of Y, then .
Remark 7. ([26], Lemma 4.4) Let be a zero-dimensional scheme such that and is linearly dependent. Since ν is an embedding, and is a line. Since is scheme-theoretically cut out by quadrics and , then . Since the only linear subspaces contained in the Segre variety X are the one contained in a fiber of one of its k rulings, there is of such that for all , is an embedding and is a line. Remark 8. Take with , and B curvilinear. The point q has border rank ([27], Proposition 1.1 and Theorem 1.2) and the minimal multiprojective space with contains all curvilinear degree b schemes such that and are contained in ([28], Theorem 2.5). Example 1. Let be a zero-dimensional scheme such that , W spans , W is linearly dependent, but all proper subschemes are linearly independent. Obviously . Since and for all lines , and either W is the complete intersection of 2 conics or for exactly one line L and there is such that all is of the form with . In both cases there are non-curvilinear W’s. In the latter case the latter case and W is the union of the fat point and some .
Example 2. Let be a degree 4 scheme. Since , if and only if there is containing W. If is linearly independent for all , then D is unique, because in this case.
Proposition 1. Let be an arrow such that Y is the minimal multiprojective space. Set . Fix . There is a zero-dimensional scheme such that , , and if and only if one of the following cases occur:
- 1 .
, ;
- 2 .
, ;
- 3 .
, .
- (a)
In case (1) we may take (with ), and reduced or and Z an arrow.
- (b)
In case (2) is the complete intersection of two conics in the plane .
- (c)
In case (3) there are reduced Z (parametrized by a plane minus a line) and arrows Z (parametrized by a line minus the point corresponding to v).
Proof. Cases (1), (2) and (3) occur and in each case the possible schemes Z described in (a), (b) and (c) are the schemes which occur (Examples 1 and 2).
Now we prove the “only if” part. Set . Thus, . Since , . First assume . We are in the case (1) by Example 1. Now assume and hence . Theorem 7 gives that we are in one of these cases (1), (2) and (3). □
Remark 9. Proposition 1 describes all degree 4 schemes , W not reduced and not containing a connected component of degree at least 3, such that Y is the minimal multiprojective space containing W, is linearly dependent and all proper subschemes of are linearly independent. For the case W reduced, see [29] and/or [30]. For an arbitrary W of degree 4, see [31]. In [26,29,30,31] there are related results obtained under assumptions with minor differences. For instance in [26] we assume that q has rank 3 and that Y is the minimal multiprojective space such that . Here we do not assume that is the minimal Segre spanning q, because it seems too restrictive. Making the assumption that is concise for q would drastically cut some proofs. Requiring that q has not cactus rank would allow the interested reader to omit Propositions 1 and 2. Remark 10. Assume .
(a) Assume for at least one i, say . Since any matrix has rank , for each there are infinitely many such that and . Now assume . Since is the tangential variety of , we obtain the existence of an arrow such that . By ([32], Ex. II.3.22(b)) we see that q is associated with at least sets S and arrows Z. (b) Assume and hence . There are with tensor rank 2 and tensors spanned by with Z connected and of degree 2. As in part (a) we obtain with and A union of an arrow and a point.
Proposition 2. Let be the union of an arrow v and a point . Assume that Y is the minimal multiprojective space containing A. Fix such that there is no with . There is a zero-dimensional scheme such that and if and only if A, Y and Z are as follows:
- 1.
and (here , say );
- 2.
, ;
- 3.
, ; ;
- 4.
, .
- (i)
In case (1) the schemes Z are as follows:
- (i1)
and Z is the point of Y such that ;
- (i2)
Z is any degree 2 subscheme.
- (ii)
In case (2) there are schemes Z formed by 2 points and schemes Z which are arrows.
- (iii)
In cases (3) and (4) there are at least schemes Z formed by 2 points and at least schemes Z which are arrows; in this case Y is not the minimal multiprojective space such that .
Proof. All listed cases are associated with some Z and the schemes Z are as described in Remarks 10 and 11. Thus, it is sufficient to prove the “only if” part.
Since and Y is the minimal multiprojective space containing A, for all i.
Set . Please note that . Since there is no such that , we have . Since Y is the minimal multiprojective space containing A, for all i. If , then . If , then and is linearly independent. Proposition 1 covers the case . Thus, we may assume , i.e., .
If , then Z is any degree 2 scheme spanning a line containing and not intersecting A. All cases with are covered by Remark 10. Thus, we may assume .
- (a)
Assume for the moment for some i, say and hence . Since Y is the minimal multiprojective space containing Y, and is linearly independent. Take and . The scheme is contained in Z.
- (a1)
Assume
. Since
, quoting ([
33], Lemma 5.1) we obtain
. Since
and
is globally generated, we obtain
and
for all
. Thus,
. Fix
. Since
and
, ([
33], Lemma 5.1) gives
. Since
and
is an embedding with linearly independent image, we obtain a contradiction.
- (a2)
Assume
, i.e., assume
. By step (a1) we may also assume
with
. Set
:=
and take
and
. Since
, we also obtain
and
. Assume that either
or
, say
. We get
. The residual exact sequence of
and ([
33], Lemma 5.1) give a contradiction because
and
is linearly independent. Assume that either
or
, say
. The residual exact sequence of
gives a contradiction. Thus,
. Using
we obtain
with
the union of
u and a point of
, call it
e. Since
,
for all
. Take
. Using
R we obtain
. Using
we obtain
. Thus,
for all
. Since
, there are
and
such that
. First assume
. Since
, ([
33], Lemma 5.1) gives
. Since
and
is linearly independent, we obtained a contradiction. Now assume
. If
(resp.
) we may take instead of
(resp.
T) a general element of its complete linear system and obtain a contradiction, because
Y is the minimal multiprojective space containing
A. Thus,
and
. One of the two divisors
T or
, say
T, contains
v. Set
. Please note that
. We obtain
, i.e.,
for all
. If
we obtain
, a contradiction. Please note that
if
Z is connected. Assume
and hence
. Since
,
. Hence
. Since
, we obtain
. Hence the residual exact sequence of
gives a contradiction.
- (b)
Assume . All cases with are listed. Thus, we assume . Let be the maximal integer such that for some and some . With no loss of generality we may assume . Set . Let be the maximal integer such that for some and some . With no loss of generality we may assume . Set . Obviously . Since Y is the minimal multiprojective space containing A, .
- (b1)
Assume
. Quoting ([
33], Lemma 5.1) we obtain a contradiction.
- (b2)
Assume
. Thus,
. If
applying ([
33], Lemma 5.1) to
we obtain a contradiction. Assume
. The definition of
gives that each
,
, is an embedding. Fix
intersecting
. Please note that
. The residual exact sequence of
and ([
33], Lemma 5.1) gives a contradiction.
- (b3)
Assume
. The definition of
gives that each
is an embedding. Take
such that
and set
. Please note that
and that
. Since
is an embedding,
, contradicting ([
33], Lemma 5.1).
- (b4)
Assume
. If
quoting ([
33], Lemma 5.1) we obtain a contradiction. Now assume
, i.e., assume
. If there are
and
with
, then quoting ([
33], Lemma 5.1) with respect to
we obtain a contradiction. Thus, we may assume
for all
. Hence
is an embedding. Set
. Since in step (b1) we excluded the case
and
, we have
and hence
. By assumption
and hence
.
- (b4.1)
Assume . Using M instead of H in the first part of step (b4) we obtain for all . Take , , containing E. We obtain . We also have . Please note that the set is a point and that is a line of the Segre variety . We obtain . Since , we obtain , i.e., Z is not reduce. Using H instead of M we obtain . Using M and , , we obtain is contained in the point , absurd.
- (b4.2)
Assume
. In the set-up of step (b4.1) we may also assume that
for all
. Set
. Let
be a degree 2 scheme. Since
is an embedding, there is
such that
is an embedding. Let
be an element of
containing a point of
. Fix
and set
. Since
,
. By ([
33], Lemma 5.1) we first obtain
and then
for all
. Either
E is the union of 3 points, say
, or the union of a point
a and an arrow
z.
- (b4.2.1)
Assume
. Thus,
. Take
. Taking
containing
c we obtain
for all
. Taking
containing
b we obtain
for all
. Thus,
for all
. Varying
i we obtain
for all
. In this case
is a line contained in the
-ruling of the Segre
. Thus,
is linearly dependent. Since
, Proposition 1 gives
. Since
and
for any
, we obtain
. Since
is linearly independent,
. Thus,
is contained in the plane,
, which also contains the line
. Since
for any
,
for any
and
,
for all
. Thus,
(scheme-theoretically). Since
contains no plane
. Since
and
is scheme-theoretically cut out by quadrics,
is the union of 2 lines. Since
Y is the minimal multiprojective space containing
A, we obtain
(([
29], Proposition 5.2) or ([
30], Proposition 1.1) or Theorem 1), a contradiction.
- (b4.2.2)
Assume
and set
. Take
. We obtain
for all
. Varying
i we obtain
for all
. Take
, but call
the integer associated with this degree 2 scheme. We obtain
for all
. Thus,
. Thus,
is a line contained in the
w-th ruling of
. If
we conclude as in step (b4.2.1). Assume
. Either
or
Z is connected and
. Take
and
. By construction
. Thus,
, contradicting ([
33], Lemma 5.1).
□
Remark 11. Take and hence . A general is contained in for sets such that , unions A of an arrow and a point and for exactly one set such that . A general q in the tangential variety of is contained in for sets such that , unions A of an arrow and a point and exactly one arrow, but no set A with cardinality 2.
Example 3. Take and take such that and there is with . This case covers cases (4) and (5) of ([26], Theorem 7.1), case (4) being the case H irreducible, while case (5) being the case H reducible. If H is irreducible (resp. it is reducible), then q is in the linear span of (resp. ) subsets of with cardinality 3. For many we have for some union A of an arrow and a point. Lemma 3. Let be a planar double point. Set and write with v and w arrows with o as their reduction. Fix such that and set and . Assume that Y is the minimal multiprojective space containing , that contains a point q and that for any . Then and if .
Proof. By assumption . Thus, there is an arrow such that . Proposition 1 shows that the minimal multiprojective space containing is either a projective space or . If , then the lemma is true. Assume . Since , we obtain , a contradiction. □
Example 4. Assume . We do the construction for the first 2 positive integers, but the case in which we take any two distinct elements of is similar. Fix arrows such that and is the minimal multiprojective space containing , but that is a plane. Thus, is a single point. Fix and . Set with embedded in the first two factors of Y. Fix , , such that for all i. Set . Thus, is a single point, . Fix and with the restriction that if and if . Set . Please note that is the line spanned by and . Please note that Y is the minimal multiprojective space containing if and only if one of the following conditions holds:
- 1.
is the minimal multiprojective space containing .
- 2.
is not the minimal multiprojective space containing , i.e., for exactly one ; in this case we require and .
Example 5. The construction done in Example 4 works if instead of the arrow we take 2 distinct points of .
Example 6. Assume . We do the construction for the first 2 positive integers, but the case in which we take any two distinct elements of is similar. Set and fix . Let L and be the elements of and containing a. Let (resp. ) be the arrow of L (resp. ) containing a. Fix and . Set with embedded in the first two factors of Y. Fix , , such that for all i. Set and . Fix and with the restriction that if and if . Set . Please note that is the line spanned by and and it is not contained in . We obtain that a general has rank 2. Please note that Y is the minimal multiprojective space containing if and only if and .
Example 7. Take either or , . Fix and take an arrow such that and for all . Please note that with . Fix such that for . If assume . Set . Please note that Y is the minimal multiprojective space containing A. Each has tensor rank at most 2, because there is such that .
Example 8. Take either or , . Fix such that and for all . Take an arrow such that and for all . Please note that with . If assume . Set . Please note that Y is the minimal multiprojective space containing A and that for all .
Lemma 4. Take , . Theorem 3 is true for Y and all B such that and .
Proof. Assume the existence of
such that
B is either the union of an arrow and a point or the union of 3 distinct points,
and
. Set
. By Propositions 1 and 2 we may assume that
irredundantly spans
q. Since
, there is
and
such that
contains
B. Since
and
Y is the minimal multiprojective space containing
A,
. Thus, the residual exact sequence of
H and ([
33], Lemma 5.1) give
. This is true for all
whose union contains
B. Since
Y is the minimal multiprojective space containing
A,
spans
. Thus,
. We obtain that
spans
. Thus, we obtain that the lines of
spanned by degree 2 subschemes of
A and
B are the same. We also obtain that
is an embedding and then we get that
is an embedding. Since
, ([
33], Lemma 5.1) and the residual exact sequence of
gives
. Please note that
is a degree 2 reduced scheme, one of its 2 points being in
, while the other one is an element of
. Since
and
are embeddings with the same set-theoretic image,
B is the union of an arrow and a point, say
with
. Taking
containing
we obtain
for all
. Using
and
in the same way we obtain
for all
, contradicting the assumption
. □
Lemma 5. Take , . Theorem 3 is true for Y and all B such that and .
Proof. Assume the existence of
such that
B is either the union of an arrow and a point or the union of 3 distinct points,
and
. Set
. By Propositions 1 and 2 we may assume that
irredundantly spans
q. Fix
. Mimicking the proof of Lemma 4 using
and
instead of
and
and doing it for all
i we obtain that
B is the union of a point
and an arrow
that each
and
are embeddings with the same images, that
spans
and that
and
for all
. Set
. We proved that
for all
. Take
containing
and
. Please note that
. Since
, ([
33], Lemma 5.1) and the residual exact sequence of
give
. Thus,
, a contradiction. □
Remark 12. Assume that Y is the minimal multiprojective space containing , , and that . Fix and assume that is not an embedding. Since Y is the minimal multiprojective space containing A, we obtain , for all and that is an embedding. Since is an embedding, we obtain that is an embedding for all .
- (a)
Assume . In this case . We obtain that is a line contained in Y. Thus, all points of have tensor rank at most 2. This is Example 7.
- (b)
Assume . In this case for all . Thus, is a line contained in the i-th ruling of the Segre , say . For each we have . In particular uniqueness fails for each q irredundantly spanned by . This case obviously occurs both with and with . This is Example 8.
- (c)
Fix a set such that and Y is the minimal multiprojective space containing Y. Assume the existence of such that is not injective. By ([26], Remark 1.10) each point of has tensor rank at most 2.
Since is linearly independent, there is such that is linearly dependent and is minimal with this property. If we obtain From now on
Lemma 6. Take , . Theorem 3 is true for Y and all B such that and .
Proof. Assume the existence of such that B is either the union of an arrow and a point or the union of 3 distinct points, and . Set . By Propositions 1 and 2 we may assume that irredundantly spans q. If B contains an arrow, then . If B is formed by 3 distinct points, then .
- (a)
Assume for the moment that
Y is not the minimal multiprojective space containing
B. Thus, there is
and
such that
and hence
. Since
, ([
33], Lemma 5.1) gives
. Since
, either
is not an embedding or
and
for
indices
j, contradicting the minimality of
Y. Thus,
Y is the minimal multiprojective space containing
B. If there is
such that
is not an embedding, we apply Remark 12 to
B.
- (b)
By assumption we are not as in Example 7 or 8 for A or B. Thus, we may assume that all and all are embeddings. There are 4 degree 2 schemes formed by one point of and one point of . Since , there is such that is an embedding. With no loss of generality we may assume .
- (b1)
Assume that
is an embedding. Take
containing a reduced connected component
a of
B. Please note that
. Since
and
Y is the minimal multiprojective space containing
Y,
is the minimal multiprojective space containing
and
is linearly independent by Remark 7. Since
is an embedding,
is an embedding. Thus,
. Since
and
is an embedding, ([
33], Lemma 5.1) gives
. Since
and
Y is the minimal multiprojective space containing
Y,
is linearly independent by ([
26], Lemma 4.4). Let
the minimal subscheme of
containing
A such that
. If
(resp.
), Proposition 1 (resp. Proposition 2) gives that
depends on at most 2 (resp. 3) factors of
. Thus,
A depends on at most 4 factors of
Y, a contradiction.
- (b2)
Assume that
is not an embedding. If there is
such that
and
, then we may repeat the proof of step (b1). Since
Y is the minimal multiprojective space containing
B by step (a), we conclude if there is
such that
and
. Thus, we may assume
. Since
and
, ([
33], Lemma 5.1) gives
. Since
is an embedding,
is an embedding. Thus,
. Let
be a minimal subscheme of such that
. If
, then Proposition 2 gives that
depends on at most 3 factors of
Y, one of them being the first one.
□
Example 9. Assume , that Y is the minimal multiprojective space containing A. Fix with tensor rank . If q has tensor rank 3, then its tensor rank is evinced by sets with , because is defective ([34,35,36]). Observation 1: Assume that is an embedding for all . There are morphisms such that and . Thus, we obtain different morphism such that and and the linear spans of the images of by these morphisms covert= the projective tangent space except its 4 hyperplanes tangents to the hypersurfaces . Hence for some f. Thus, is contained in the rational normal curve of the 4-dimensional space . The point q has cactus rank with respect to C. By Sylvester’s theorem and the C-rank is achieved by subsets of C with cardinality 3.
Now assume that is not an embedding for some i, say for . We extend taking instead of the open subset U of the irreducible component of the Hilbert scheme of containing and formed by degree 3 linearly independent schemes. Please note that and that A is limit of elements such that each is a union of an arrow and a point and is an embedding for all and all . Use Observation 1 and ([32], Ex. II.3.22) to see that in this case q is spanned by infinitely many . Since A has 2 connected components, each nearby has at least 2 connected components, i.e., it is the union of either 3 points or a point and an arrow. Proof of Theorem 3. If , then we are in the set-up of Propositions 1 and 2. Thus, we may assume .
If we use Remark 10. Thus, we may assume .
If we use Proposition 2 and the examples listed in case (7) of the theorem. Thus, we may assume .
If we use Lemma 4.
If and we use Lemma 5.
Example 9 consider the case .
Lemma 6 exclude the case , . □
6. Speculations on the Higher Derivatives and Uniqueness
Fix integral and non-degenerate varieties and . Set and . Fix a finite set , , and . Set . Suppose that (resp. ), i.e., assume that S is in the k-Terracini locus of S (resp. the pair is in the Terracini locus of the join of X and W). Thus, the differential of a certain map, call it f, is not injective at S (resp. at ). It is natural to ask if this is due that the fact that the fiber of f containing S (resp. ) has a positive-dimensional component passing through S (resp. ). Easy examples shows that this is not always the case (Examples 10 and 11).
Example 10. Fix integer . We claim the existence of a smooth, rational and non-degenerate curve and such that , and , i.e., and . In this case the fiber of the abstract 2-secant map of X at is finite. Fix an integer . Let be the rational normal curve. Fix such that . Since C is a rational normal curve and , the degree 6 scheme is linearly independent. Thus, . Fix a general line and a general linear subspace such that and . Since and M is general, .
Claim 1: and the linear projection from M induces an isomorphism between C and the degree d curve .
Proof of Claim 1: Since C is smooth, each point of is contained in a tangent line or a secant line of C. Since is a rational normal curve, every closed subscheme of C of degree at most is linearly independent. Thus, the assumption implies . Since L is general in and M has codimension at least 3 in , and hence and induces an isomorphism between C and the degree d non-degenerate curve X. □
Claim 1 shows that and . Since , the fibers of the map have dimension 0.
Example 11. Fix integer . We claim the existence of a smooth, rational and non-degenerate curve and such that , is a point not in X and , i.e., and . In this case the fiber of the abstract 2-secant map of X at is finite. Fix an integer . Let be the rational normal curve. We adapt Example 10. Instead of L we take a general and take as M a general linear subspace of codimension containing c.
Fix an integral and non-degenerate
n-dimensional variety
and a positive integer
s such that
. Let
denote the set of all
with
X-rank
r. Let
denote the set of all
which are in the linear span of a unique subset of
X with cardinality. The sets
and
are constructible and the first one contains a non-empty open subset of
. Assume that
has the expected dimension,
, to hope to have
. There are criterion which guarantee that
is dense in
, (for any
X weak nondefectivity and tangential nondefectivity ([
10,
11,
12]), for tensors and partially symmetric tensors the famous Kruskal criterion and its modifications; up to now for tensors the best results are in [
37]).
Question 4. Assume dense in . Describe in specific cases the non-uniqueness set .
There are examples with
,
containing a general
,
and
very small (linearly normal embeddings of elliptic curves). If
, say
with
and
,
q is in the linear span of all
,
, but none of these sets irredundantly spans
q. For tensors a general tensor of each rank
s is irredundantly spanned by sets of
t points if
in infinitely many ways ([
38], Theorem 3.8). The following example with
z any positive integer shows that sometimes above its rank
a point
q may be irredundantly spanned by a unique set of cardinality
.
Example 12. We first construct a smooth curve , , and such that there is a unique with and . Let be a rational normal curve of degree . Fix 3 distinct points a, b and c of C and let o be a general element of . Let denote the linear projection from o. By Sylvester theorem (or because any subscheme of C of degree is linearly independent) there is no such that and . Thus, and induces an embedding of C into . Set , and . Thus, and the set S irredundantly spans q. Assume the existence of a set such that and irredundantly spans q, i.e., and . Set and . Since ℓ is an embedding, . Since , . Thus, contradicting the fact that the lines and meets at q. Now we fix an integer and modify the previous example to obtain one for z. Fix an integer . Let be a rational normal curve. Fix general points and a general . Since , any closed subscheme of C of degree is linearly independent. Let denote the linear projection from o. The map induces an embedding . Set . By construction the -dimensional space meets the z-dimensional space at a unique point, q. Since any points of C are linearly independent, we first see that q has X-rank z, then that is the unique set evincing the X-rank of q and then that is the unique subset of X with cardinality irredundantly spanning q.