On 12/15/17 12:14 AM, JanPB wrote:
> Symbols like "dx", "ds^2", etc. denote tensors, not matrices.
Yes. But some people around here, who learned from different textbooks, could
well be confused by this. So I'll explain the three COMPLETELY DIFFERENT
meanings of these symbols. In a textbook they would not be confused, because
tensors would be typeset as bold; here we do not have that ability.
Consider this equation, in the context of GR and a specific 4-D manifold:
ds^2 = g_ij dx^i dx^j (1)
Interpretation A:
The {x^k} are coordinates on a region of the manifold, which means that each
point in the region corresponds to a 4-tuple of coordinate values (x^0, x^1,
x^2, x^3); i, j, and k are indexes which range from 0 to 3. ds is an
infinitesimal real number, ds^2 is its square, and the {dx^k} are real numbers
representing infinitesimal displacements along the corresponding coordinate
axes. The {g_ij} are a set of real functions of the {x^k}, called the COMPONENTS
of the metric tensor. Equation (1) is called "the line element", because it
expresses the notion that for an infinitesimal line with coordinate
displacements {dx^k}, its length will be ds. The metric tensor g is:
g = g_ij e^i e^j
where the {e^k} are the basis co-vector fields of the coordinate system {x^k};
they are labeled with an upper (contravariant) index, but each is a covariant
vector.
Interpretation B:
The {x^k} are coordinates on a region of a 4-D manifold, which means they are
real functions on the manifold (i.e. each of them maps each point of the region
to a real number, in a systematic and continuous manner). Each of the {dx^k} is
a ONE-FORM, because here d is the exterior derivative (a one-form is a rank-1
covariant tensor function on the manifold). The {g_ij} are a set of real
functions on the manifold, called the components of the metric tensor. Equation
(1) is called "the metric tensor" because it defines ds^2 as a symmetric rank-2
tensor; it is obviously a function on the manifold (aka a tensor field on the
manifold). The metric tensor g is:
g = ds^2
Interpretation C:
The {x^k} are coordinates on a region of a 4-D manifold, which means they are
real functions on the manifold (i.e. each of them maps each point of the region
to a real number, in a systematic and continuous manner). Each of the {dx^k} is
a ONE-FORM, because here d is the exterior derivative (a one-form is a rank-1
covariant tensor function on the manifold). g_ij is the metric tensor, in
abstract-index notation (note the absence of {}, because it is a single object,
a rank-2 tensor, not a set of components). ds is an infinitesimal real number,
ds^2 is its square. The right-hand side of (1) is the contraction of three
tensors. The metric tensor is g_ij (not g, because this is abstract index notation).
Many textbooks use interpretation A; Jan is using interpretation B. Wald uses C.
Koobee Wublee uses NONE OF THESE, which is why he is so confused
and makes so many incorrect statements. He thinks the {g_ij} of A
are "the metric tensor, which is a matrix". While it is true that
the {g_ij} form a matrix, they are not at all a tensor -- see above
for what the metric tensor actually is. Some archaic textbooks also
get this wrong and sowed much confusion. In particular:
Weinberg, _Gravitation_and_Cosmology_ **AVOID**
Eisenhardt, _Riemannian_Geometry_ **AVOID**
Here are some books that get it right:
Misner, Thorne, and Wheeler, _Gravitation_
Hawking and Ellis, _The_Large_Scale_Structure_of_Space-Time_
Frankel, _The_Geometry_of_Physics_
Baez and Muniain, _Gauge_Fields,_Knots,_and_Gravity_
Wald, _General_Relativity_
Tom Roberts