Namespaces
Variants
Actions

Difference between revisions of "Seifert matrix"

From Encyclopedia of Mathematics
Jump to: navigation, search
m (tex encoded by computer)
(details)
Line 11: Line 11:
 
{{TEX|done}}
 
{{TEX|done}}
  
A matrix associated with knots and links in order to investigate their topological properties by algebraic methods (cf. [[Knot theory|Knot theory]]). Named after H. Seifert [[#References|[1]]], who applied the construction to obtain algebraic invariants of one-dimensional knots in  $  S  ^ {3} $.  
+
A matrix associated with knots and links in order to investigate their topological properties by algebraic methods (cf. [[Knot theory]]). Named after H. Seifert [[#References|[1]]], who applied the construction to obtain algebraic invariants of one-dimensional knots in  $  S  ^ {3} $.  
 
Let  $  L = ( S  ^ {n+} 2 , l  ^ {n} ) $
 
Let  $  L = ( S  ^ {n+} 2 , l  ^ {n} ) $
 
be an  $  n $-
 
be an  $  n $-
 
dimensional  $  m $-
 
dimensional  $  m $-
component [[Link|link]], i.e. a pair consisting of an oriented sphere  $  S  ^ {n+} 2 $
+
component [[link]], i.e. a pair consisting of an oriented sphere  $  S  ^ {n+} 2 $
 
and a differentiable or piecewise-linear oriented submanifold  $  l  ^ {n} $
 
and a differentiable or piecewise-linear oriented submanifold  $  l  ^ {n} $
 
of this sphere which is homeomorphic to the disconnected union of  $  m $
 
of this sphere which is homeomorphic to the disconnected union of  $  m $
Line 55: Line 55:
 
$$
 
$$
  
where the right-hand side is the [[Intersection index (in homology)|intersection index (in homology)]] of the classes  $  z _ {1} $
+
where the right-hand side is the [[Intersection index (in homology)|intersection index]] of the classes  $  z _ {1} $
 
and  $  z _ {2} $
 
and  $  z _ {2} $
 
on  $  V $.
 
on  $  V $.
Line 65: Line 65:
 
with integer entries is called the Seifert matrix of  $  L $.  
 
with integer entries is called the Seifert matrix of  $  L $.  
 
The Seifert matrix of any  $  ( 2 q - 1 ) $-
 
The Seifert matrix of any  $  ( 2 q - 1 ) $-
dimensional knot has the following property: The matrix  $  A = ( - 1 )  ^ {q} A ^ \prime $
+
dimensional knot has the following property: The matrix  $  A = ( - 1 )  ^ {q} A^t $
 
is unimodular (cf. [[Unimodular matrix|Unimodular matrix]]), and for  $  q = 2 $
 
is unimodular (cf. [[Unimodular matrix|Unimodular matrix]]), and for  $  q = 2 $
the [[Signature|signature]] of the matrix  $ A + A ^ \prime  $
+
the [[signature]] of the matrix  $A + A^t$
is divisible by $ 16 $(
+
is divisible by $16$ ($A^t$ is the transpose of $A$).  
$ A ^ \prime  $
 
is the transpose of $ A $).  
 
 
Any square matrix  $  A $
 
Any square matrix  $  A $
 
with integer entries is the Seifert matrix of some  $  ( 2 q - 1 ) $-
 
with integer entries is the Seifert matrix of some  $  ( 2 q - 1 ) $-
 
dimensional knot if  $  q \neq 2 $,  
 
dimensional knot if  $  q \neq 2 $,  
and the matrix  $  A + ( - 1 )  ^ {q} A ^ \prime $
+
and the matrix  $  A + ( - 1 )  ^ {q} A^t $
 
is unimodular.
 
is unimodular.
  
Line 106: Line 104:
 
while  $  A $
 
while  $  A $
 
itself is called an elementary reduction of its elementary expansions. Two square matrices are said to be  $  S $-
 
itself is called an elementary reduction of its elementary expansions. Two square matrices are said to be  $  S $-
equivalent if one can be derived from the other via elementary reductions, elementary expansions and unimodular congruences (i.e. transformations  $  A \rightarrow P ^ \prime A P $,  
+
equivalent if one can be derived from the other via elementary reductions, elementary expansions and unimodular congruences (i.e. transformations  $  A \rightarrow P^t A P $,  
 
where  $  P $
 
where  $  P $
 
is a unimodular matrix). For higher-dimensional knots  $  ( m = 1 ) $
 
is a unimodular matrix). For higher-dimensional knots  $  ( m = 1 ) $
Line 117: Line 115:
 
module  $  H _ {q} \widetilde{X}  $,  
 
module  $  H _ {q} \widetilde{X}  $,  
 
where  $  \widetilde{X}  $
 
where  $  \widetilde{X}  $
is an infinite cyclic covering of the complement of the knot. The polynomial matrix  $  t A + ( - 1 )  ^ {q} A  ^ \prime $
+
is an infinite cyclic covering of the complement of the knot. The polynomial matrix  $  t A + ( - 1 )  ^ {q} A  ^t $
 
is the Alexander matrix (see [[Alexander invariants|Alexander invariants]]) of the module  $  H _ {q} \widetilde{X}  $.  
 
is the Alexander matrix (see [[Alexander invariants|Alexander invariants]]) of the module  $  H _ {q} \widetilde{X}  $.  
 
The Seifert matrix also determines the  $  q $-
 
The Seifert matrix also determines the  $  q $-
 
dimensional homology and the linking coefficients in the cyclic coverings of the sphere  $  S  ^ {2q+} 1 $
 
dimensional homology and the linking coefficients in the cyclic coverings of the sphere  $  S  ^ {2q+} 1 $
 
that ramify over the link.
 
that ramify over the link.
 +
 +
====Comments====
 +
For a description of the Seifert manifold in the case $n = 1$,
 +
i.e. the Seifert surface of a link, cf. [[Knot and link diagrams]].
  
 
====References====
 
====References====
<table><TR><TD valign="top">[1]</TD> <TD valign="top">  H. Seifert,  "Ueber das Geschlecht von Knoten"  ''Math. Ann.'' , '''110'''  (1934)  pp. 571–592</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top">  R.H. Crowell,  R.H. Fox,  "Introduction to knot theory" , Ginn  (1963)</TD></TR><TR><TD valign="top">[3]</TD> <TD valign="top">  J. Levine,  "Polynomial invariants of knots of codimension two"  ''Ann. of Math.'' , '''84'''  (1966)  pp. 537–554</TD></TR><TR><TD valign="top">[4]</TD> <TD valign="top">  J. Levine,  "An algebraic classification of some knots of codimension two"  ''Comment. Math. Helv.'' , '''45'''  (1970)  pp. 185–198</TD></TR></table>
+
<table>
 
+
<TR><TD valign="top">[1]</TD> <TD valign="top">  H. Seifert,  "Ueber das Geschlecht von Knoten"  ''Math. Ann.'' , '''110'''  (1934)  pp. 571–592</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top">  R.H. Crowell,  R.H. Fox,  "Introduction to knot theory" , Ginn  (1963)</TD></TR>
====Comments====
+
<TR><TD valign="top">[3]</TD> <TD valign="top">  J. Levine,  "Polynomial invariants of knots of codimension two"  ''Ann. of Math.'' , '''84'''  (1966)  pp. 537–554</TD></TR>
For a description of the Seifert manifold in the case  $  n = 1 $,
+
<TR><TD valign="top">[4]</TD> <TD valign="top">  J. Levine,  "An algebraic classification of some knots of codimension two"  ''Comment. Math. Helv.'' , '''45'''  (1970)  pp. 185–198</TD></TR>
i.e. the Seifert surface of a link, cf. [[Knot and link diagrams|Knot and link diagrams]].
+
</table>

Revision as of 09:13, 10 April 2023


A matrix associated with knots and links in order to investigate their topological properties by algebraic methods (cf. Knot theory). Named after H. Seifert [1], who applied the construction to obtain algebraic invariants of one-dimensional knots in $ S ^ {3} $. Let $ L = ( S ^ {n+} 2 , l ^ {n} ) $ be an $ n $- dimensional $ m $- component link, i.e. a pair consisting of an oriented sphere $ S ^ {n+} 2 $ and a differentiable or piecewise-linear oriented submanifold $ l ^ {n} $ of this sphere which is homeomorphic to the disconnected union of $ m $ copies of the sphere $ S ^ {n} $. There exists a compact $ ( n+ 1) $- dimensional orientable submanifold $ V $ of $ S ^ {n+} 2 $ such that $ \partial V = l $; it is known as the Seifert manifold of the link $ L $. The orientation of the Seifert manifold $ V $ is determined by the orientation of its boundary $ \partial V = l $; since the orientation of $ S ^ {n+} 2 $ is fixed, the normal bundle to $ V $ in $ S ^ {n+} 2 $ turns out to be oriented, so that one can speak of the field of positive normals to $ V $. Let $ i _ {+} : V \rightarrow Y $ be a small displacement along this field, where $ Y $ is the complement to an open tubular neighbourhood of $ V $ in $ S ^ {n+} 2 $. If $ n = 2 q - 1 $ is odd, one defines a pairing

$$ \theta : H _ {q} V \otimes H _ {q} V \rightarrow \mathbf Z , $$

associating with an element $ z _ {1} \otimes z _ {2} $ the linking coefficient of the classes $ z _ {1} \in H _ {q} V $ and $ i _ {+} * z _ {2} \in H _ {q} Y $. This $ \theta $ is known as the Seifert pairing of the link $ L $. If $ z _ {1} $ and $ z _ {2} $ are of finite order, then $ \theta ( z _ {1} \otimes z _ {2} ) = 0 $. The following formula is valid:

$$ \theta ( z _ {1} \otimes z _ {2} ) + ( - 1 ) ^ {q} \theta ( z _ {2} \otimes z _ {1} ) = z _ {1} \cdot z _ {2} , $$

where the right-hand side is the intersection index of the classes $ z _ {1} $ and $ z _ {2} $ on $ V $.

Let $ e _ {1} \dots e _ {k} $ be a basis for the free part of the group $ H _ {q} V $. The $ ( k \times k ) $- matrix $ A = \| \theta ( e _ {i} \otimes e _ {j} ) \| $ with integer entries is called the Seifert matrix of $ L $. The Seifert matrix of any $ ( 2 q - 1 ) $- dimensional knot has the following property: The matrix $ A = ( - 1 ) ^ {q} A^t $ is unimodular (cf. Unimodular matrix), and for $ q = 2 $ the signature of the matrix $A + A^t$ is divisible by $16$ ($A^t$ is the transpose of $A$). Any square matrix $ A $ with integer entries is the Seifert matrix of some $ ( 2 q - 1 ) $- dimensional knot if $ q \neq 2 $, and the matrix $ A + ( - 1 ) ^ {q} A^t $ is unimodular.

The Seifert matrix itself is not an invariant of the link $ L $; the reason is that the construction of the Seifert manifold $ V $ and the choice of the basis $ e _ {1} \dots e _ {k} $ are not unique. Matrices of the form

$$ \left \| \begin{array}{lcc} A &{} & 0 \\ \alpha & 0 & 1 \\ 0 & 1 & 0 \\ \end{array} \right \| ,\ \ \left \| \begin{array}{lll} A &\beta & 0 \\ 0 & 0 & 1 \\ {} & 0 & 0 \\ \end{array} \right \| , $$

where $ \alpha $ is a row-vector and $ \beta $ a column-vector, are known as elementary expansions of $ A $, while $ A $ itself is called an elementary reduction of its elementary expansions. Two square matrices are said to be $ S $- equivalent if one can be derived from the other via elementary reductions, elementary expansions and unimodular congruences (i.e. transformations $ A \rightarrow P^t A P $, where $ P $ is a unimodular matrix). For higher-dimensional knots $ ( m = 1 ) $ and one-dimensional links $ ( n = 1 ) $ the $ S $- equivalence class of the Seifert matrix is an invariant of the type of the link $ L $. In case $ L $ is a knot, the Seifert matrix $ A $ uniquely determines a $ \mathbf Z [ t , t ^ {-} 1 ] $- module $ H _ {q} \widetilde{X} $, where $ \widetilde{X} $ is an infinite cyclic covering of the complement of the knot. The polynomial matrix $ t A + ( - 1 ) ^ {q} A ^t $ is the Alexander matrix (see Alexander invariants) of the module $ H _ {q} \widetilde{X} $. The Seifert matrix also determines the $ q $- dimensional homology and the linking coefficients in the cyclic coverings of the sphere $ S ^ {2q+} 1 $ that ramify over the link.

Comments

For a description of the Seifert manifold in the case $n = 1$, i.e. the Seifert surface of a link, cf. Knot and link diagrams.

References

[1] H. Seifert, "Ueber das Geschlecht von Knoten" Math. Ann. , 110 (1934) pp. 571–592
[2] R.H. Crowell, R.H. Fox, "Introduction to knot theory" , Ginn (1963)
[3] J. Levine, "Polynomial invariants of knots of codimension two" Ann. of Math. , 84 (1966) pp. 537–554
[4] J. Levine, "An algebraic classification of some knots of codimension two" Comment. Math. Helv. , 45 (1970) pp. 185–198
How to Cite This Entry:
Seifert matrix. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Seifert_matrix&oldid=49578
This article was adapted from an original article by M.Sh. Farber (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article