Example : InternaSymmetri
The above examples involve transformations of spacetime ,
as well
transformations of
as
the fields . An internal
symmetry is one that
only involves a transformation of the fields
acts at
and the same
every point in spacetime .
Let
of free field theory > OuY & y * -P N.
&
he consider the example complex scalar -
2 =
- e by
This
Lagrangian has a continuous U(I) symmetry as the UCL) transformation ,
4
leaves L unchanged . We add further terms proportional to &4, and
can I would still
be invariant.
**
For infinitesimal transformations : 4- e + = Y + 1)4.
: D4 = 14 and
similarly Dr * = - i **. Furthermore , DC = 0 E FM = 0 .
: The conserved current >
-
JM
= D4
+)D4 *
=
i[ + (v) =(0mp)]
Note that OnJM = 0
gives the exactly same continuity equation that you have seen
for Klein-Gordon equation in relativistic quantum mechanics classes ,
where the
density component (i time like
component P A Jo i(NOLY *
-4 *& 4) had
.
.e = =
no
., ,
interpretation in terms of probability density But later when we will consider
.
interpret electric
quantum field theory we will correctly 30 as
charge density or
particle number density
.
Alternatively ,
we can consider two coupled real scalar fields >
-
C =
404-m-g
This system is
equivalent free complex
to the scalar
theory for g = 0
,
where
↑ =F (4 + id) and 44 = (4-i4) The ·
last term defines interactions between the
fielde by and 42 ,
and is
equivalent to
adding a
( **) term in the complex field theory
Lagrangian .
The above
Lagrangian is invariant under so(z) volations >
-
>
-
3) + ↑ sind
*- 4 ,
02d - find
This is not
surprising as 30(2) is isomorphic to U(I).
For infinitesimal transformations - 1 + 1 :
and -4- 34 ·
Therefore D& * and DK -d while D2
,
= =
,
and Fr are zero .
: converted current >
-
>M =
4(044) ↑(0M42).
-
similarly ,
one can have non-Abelian internal
symmetries . For example , if we consider
the
following Lagrangian N
of real scalar fields >
) -
-
C =
then the Lagrangian is invariant under SO(N) rotation among I fields . One can construct
the
corresponding conserved current
following the same
procedure .
Now ,
how do the above discussions transtate into quantum case ? Here we will briefly
answer this for quantum particle mechanics ,
as we still have to
get familiar with quantum
field theory
.
The discussione straight forwardly translate quantum to mechanics where variable are
promoted to operators following canonical quantization rule : E ,
SpB . FC ,
J .
So , for a
continuous
symmetry ,
we will have a conserved
quantity which is now an operator.
There will be
ordering ambiguities when
promoting & to an
operator ,
as operators do not
commute in
general We will talk about this when we encounter this in quantum field theory
.
.
classical mechanics that [Q H3p B = 0
In
,
one can show Qobeys ,
.
and Ea Q3pB ,
=
Daj ·
expected
.
The first relation tells the = 0
,
while from the second relation we can
infer the (infinitesimal) symmetry transformation by knowing the conserve
quantity , i. e
.,
the opposite to the Noethere procedure .
In quantum particle mechanics these relations translates into - ] 0 and
=
[ ] : D If you use these for translation and time translation symmetries , you
,
=
.
will get familiar relations in quantum mechanics .
For classical field theory we will get [Q HJp B= 0
, ,
.
and [4 QYpB.
,
=
D4j ·
Later we will
see what happens for quantum field theory.
↓ note on functional calcu.
In most of the QFT books you
will encounter functional
calculus ,
specifically functional derivatives F lim(F) =
+ -x]F]) during o
the discussions classical
on
field theory
.
This is due to the fact that in classical field theory ,
Lagrangian L
JarC([43 3043 27)
=
, ,
and Hamiltonian H =
/dr2 (44;3 , [51j3 24)
,
are
functionals of the fields It) (and of Outs
or JA)) that are
,
in turn , functions of
In terms of functional derivatives ,
the conjugate fields >
-
Dj = /C =t =
·
This is because
>= JC (Ei
-
,
30ui
=
Ent[(ir(24(x) + top x)3 50m4, b)3
-
,
,
24) [x = x4 is a 4-vector]
-
Jar'd (54()3 [043 zim)] , ,
Now , ((34j() + E8(-2)] &Outjbe)3 zin) , ,
=
L(54; (3 ,
[Ou4; (2)3 ,
zen) + E@-x) Ol)
Therefore , )
Similarly t=
> ,
one can show -
Similarly ,
a mathematically vigorous definition of Poisson bracket would be > -
EF G3
,
=
Jirit)it)]
However ,
we don't have to make our
life complicated by using gymnastics of
functional derivatives treat of
.
If we just as an index like ; (as we have done in
our discussions) ,
we can handle functional derivatives in the same
way as standard derivatives.
mathematician friends will not there will few (minor)
Our be
happy and by a
subtleties , but as
physicists we will be just fine.
So ,
for our
purpose >t =) ot) Jo C (4it)
-
=
,
On 4 fit), 2
=>
ja
= O = av)
↓
This is the subtlety and
.
non-rigorous part
Similarly ,
we do the same for Poisson .
brackets