On Fermi systems with strong forward scattering in d spatial dimensions
Extracted source text ยท arXiv cond-mat/0004451
\documentstyle[prl,aps,psfig]{revtex}
%\documentstyle[aps,12pt,psfig]{revtex}
%
\begin{document}
\draft
\title{On Fermi systems with strong forward scattering in $d$ spatial
dimensions
}
\author{\sc Behnam Farid}
\address{Cavendish Laboratory, Department of Physics,
University of Cambridge,\\
Cambridge CB3 0HE, United Kingdom
\thanks{Electronic address: bf10000@phy.cam.ac.uk}\\
and Max-Planck-Institut f\"ur Festk\"orperforschung,
Heisenbergstra\ss e 1,\\
70569 Stuttgart, Federal Republic of Germany
\thanks{Electronic address: farid@audrey.mpi-stuttgart.mpg.de}\\
}
\date{\today }
\maketitle
\vspace{0.66cm}
\begin{abstract}
{\sf
We consider a specific generalisation to spatial dimensions $d$ greater
than one of a formalism based upon conservation laws and the associated
Ward identities, that exactly solves the one-dimensional Luttinger model,
and expose its inadequacies for $d > 1$. We conclude that findings
arrived at through application of this generalised formalism concerning
systems of fermions with strong forward scattering, in particular that
$d=1$ would be the cross-over dimension from Luttinger- to Fermi-liquid
state, are open to question.
}
\end{abstract}
\vspace{0.33cm}
%__________________________________ Definitions: _____
\def\Bf#1{\mbox{\boldmath{$#1$}}}
\def\DL{Dzyaloshinski\v{i}-Larkin}
%_____________________________________________________
%
%_________
\vskip 2.8cm
\noindent\underline{\sf To appear in:
{\sl Philosophical Magazine, Part B, Vol.~{\Bf 80}, No.~9 (2000)} }
\vskip 0.25cm
%_________
\maketitle
%
%\vfill
%\pagebreak
% Sec. I:
\section{Introduction}
{\Large\bf T}he question with regard to the existence of non-Fermi
liquid (NFL) metallic states in spatial dimensions $d$ greater than one,
specifically in $d=2$, has been subject of intensive investigations in
the course of the past several years, following the observation that
the normal-state properties of doped cuprate superconducting compounds
markedly differ from those of conventional metals which are accurately
described within the framework of the Fermi-liquid (FL) theory (Anderson
1988,1989,1990a,b,1997, Varma, Littlewood, Schmitt-Rink, Abrahams and
Ruckenstein 1989, 1990, Littlewood and Varma 1991). Since according to
a celebrated theorem due to Luttinger (1961), validity of the many-body
perturbation theory to all orders would imply that metallic states of
fermions in spatial dimensions greater than one would be FLs,
\footnote{\label{f1}
Strictly speaking, this theorem has direct bearing on cases where
fermions interact through a {\sl short-range} two-body potential.}
failure of the FL picture in the case of the doped cuprates has been
considered to signal breakdown of the many-body perturbation theory as
applied to these materials. Indeed, in the course of the recent years the
question with regard to breakdown of the many-body perturbation theory
has been a central aspect in debates concerning the above-mentioned
failure of the FL phenomenology in describing the normal-state properties
of the doped cuprate compounds (Anderson 1990b, 1993, Engelbrecht and
Randeria 1990, Stamp 1993) (see also Engelbrecht and Randeria (1991),
Anderson (1991)). Consequently, in this period a number of non-perturbative
methods have been put forward and applied to models of interacting fermions
with the aim of determining whether or not NFL states are viable in $d > 1$.
Elsewhere (Farid 1999a,b) we have presented accounts of these developments
and therefore refrain from repetition here. For our present considerations
it is relevant to mention that the existence of NFL metallic states in
realistic {\sl microscopic} models of interacting fermions in $d > 1$ has
remained elusive to this date.
In view of the above-indicated apparent discrepancy that prevails between
theoretical predictions and experimental observations, we have undertaken
to subject a number of aspects that are crucial to establishing feasibility
or otherwise of NFL metallic states in $d > 1$ to an independent
investigation. The present work is the third report of our investigations
on this subject. Our first work (Farid 1999a) concerned a critical
analysis of the above-indicated Luttinger (1961) theorem. We have
demonstrated (Farid 1999a) that Luttinger's (1961) proof of this theorem
involves an implicit assumption which is {\sl strictly} specific to FL
metallic states, implying that the Luttinger (1961) theorem far from
excluding the possibility of the existence of NFL metallic states in
$d > 1$ (for cases where the many-body perturbation theory is valid to
{\sl all} orders), in fact amounts to a {\sl statement of consistency}
of the FL states with the principles of the theory of interacting
fermions as expressed in terms of the many-body perturbation theory.
Our second work (Farid 1999b) concerned an in-depth analysis of the
non-perturbative bosonisation scheme due to Haldane (1992) for metallic
states in $d \ge 1$ as applied to a model of interacting fermions in $d=2$,
originally put forward by Houghton and Marston (1993) and subsequently
extensively studied by Houghton, Kwon and Marston (1994), Houghton, Kwon,
Marston and Shankar (1994) and Kwon, Houghton and Marston (1995). We have
demonstrated failure of the Haldane (1992) bosonisation scheme in $d > 1$
and established that this scheme invariably predicts the metallic state of
the model under consideration to be an {\sl unconventional} FL,
irrespective of the nature ({\sl and} strength, provided it be
non-vanishing) of the fermion-fermion interaction function, that is
whether this be short-range, long-range Coulomb or super-long range.
The latter result is in particular at variance with an earlier finding
by Bares and Wen (1993),
\footnote{\label{f2}
It is also at variance with the findings by Houghton, Kwon, Marston and
Shankar (1994) and Kwon, Houghton and Marston (1995). We have however
shown (Farid 1999b) that this discrepancy is due to use of certain
invalid approximations by the latter authors. }
indicating the metallic state of the system for super-long-range
interactions to be a NFL. We have traced back the shortcoming of the
Haldane (1992) bosonisation scheme in $d > 1$ to the neglect of some
terms in the commutation relation concerning some current operators; in
particular, in the case of metallic systems with a sufficiently small (or
vanishing) value for the jump $Z_{{\bf k}_F}$ in the momentum distribution
function at the Fermi momentum, the neglected terms are more relevant than
the sole term that is taken into account, namely the {\sl quantum anomaly}.
\footnote{\label{f3}
For completeness, we mention that by solely taking into account the
{\sl quantum anomaly}, the mentioned current operators form a Kac-Moody
algebra (see, e.g., Goddard and Olive 1986).}
We have also shown that under {\sl all} conditions the Fermi energy of the
interacting system within the Haldane (1992) bosonisation scheme coincides
with that of the non-interacting system. This unambiguously implies that
{\sl even} for weakly-interacting metallic states, for which $Z_{{\bf k}_F}$
is close to unity, neglect of the indicated terms in the mentioned
commutation relation for current operators is {\sl not} justified.
The exactly solvable one-dimensional Luttinger model (Luttinger 1963,
Mattis and Lieb 1965) has been dealt with by three distinct methods. One
is based on the technique of bosonisation, of which the Haldane (1992)
scheme is a generalisation to higher dimensions.
\footnote{\label{f4}
The bosonisation method for calculation of correlation functions is due
to Luther (1979) which derives from works by Luther and Peschel (1974),
Mattis and Lieb (1965), Tomonaga (1950) and Bloch (1933, 1934). Parallel
developments involving similar views as in the indicated works are due
to Jordan (1935, 1936a,b), Kr\"onig (1935a,b,c), Skyrme (1958, 1959, 1961a,b)
and Coleman (1975). For reviews see Stone (1994) and Kopietz (1997).}
The second, due to Everts and Schulz (1974), is based on the solution of
the equation of motion for the Green function which derives from an exact
decoupling of the hierarchy of the Green functions. The third approach,
due to Dzyaloshinski\v{i} and Larkin (DL) (1974),
\footnote{\label{f5}
For a detailed exposition of this approach see (Bohr 1981).}
exploits the conservation laws and the associated Ward identities. This
scheme has been generalised to $d > 1$ and reported in a series of articles
by Di Castro and Metzner (1991), Metzner and Di Castro (1993), Castellani,
Di Castro and Metzner (1994), Castellani and Di Castro (1994) and
extensively reviewed by Metzner, Castellani and Di Castro (MCDC) (1998).
It is this scheme that we subject to a critical analysis in the present
work. In our considerations that follow, we shall almost exclusively
refer to the last review article without thereby implying that the cited
subject matters would have their origin in this work.
Contrary to case of $d=1$, associated with {\sl linear} energy dispersions
for the non-interacting left- and right-movers (that is the case of the
one-dimensional Luttinger (1963) model), in cases of $d > 1$ the generalised
DL approach is {\sl not} exact; MCDC (1998) assert, however, that for
Fermi systems the generalised approach is {\sl asymptotically} exact in
the limit of strong forward scattering, that is in the limit of zero
momentum transfer ${\bf q}$ for fermions in a narrow band of width
$\lambda$ ($\Lambda$ in the authors' notation) encompassing the Fermi
surface. {\sl In this work we demonstrate this not to be the case.} We
further support this finding by referring to the similarity between the
expressions for the single-particle Green function within the framework of
the generalised DL (MCDC 1998) scheme and the Haldane (1992) bosonisation
approach and the fact that, as we have indicated above, the latter is {\sl
not} adequate for $d > 1$ (Farid 1999b). We also establish that the
rigorous formalism for systems of interacting fermions, even in the limit
of strong forward scattering, is {\sl not} amenable to treatment by the
Wilson-Fisher (1972) (Wilson 1973) technique of dimensional continuation.
The organisation of this work is as follows. In \S~II we present an outline
of the generalised DL formalism (MCDC 1998) which will facilitate our
discussions in this paper. Anticipating the results that we present in
\S~III, here we indicate that the conventional process of dimensional
continuation is {\sl not} applicable to a general formalism concerning
systems of {\sl interacting} fermions, not even in the limit where solely
forward scattering processes are taken into account. In \S~III we analyse
two crucial steps that are involved in the derivation of the generalised
DL formalism and explicitly demonstrate their inadequacies in $d > 1$;
in \S~III.A we deal with a relationship between the charge and the current
part of a vertex function that has been held to be asymptotically correct
in the limit of strong forward-scattering and show that its validity in
$d > 1$ is {\sl strictly} limited to the trivial case of non-interacting
fermions; in \S~III.B we show the invalidity of a second relationship
between two irreducible vertex functions, also held to be asymptotically
correct in $d > 1$ in the limit of strong forward-scattering. In \S~IV
we explicitly demonstrate how the singular behaviour of the momentum
distribution function, in particular that pertaining to the ground-state
of the system, at the Fermi momentum, renders the mathematical
manipulations that are central to the derivation of the generalised DL
method in $d > 1$ (MCDC 1998) inapplicable. Here we further show how
similar manipulations, combined with a complete disregard of the errors
associated with these, in $d=1$ (specifically in dealing with the
one-dimensional Luttinger model) give rise to results which coincide with
the exact results, thus misleadingly suggesting their general applicability
even in $d > 1$. We finally present a summary and concluding remarks of
this work in \S~V. We devote an Appendix which follows the main text to
a brief exposition of the Wilson-Fisher (Wilson and Fisher 1972, Wilson
1973) technique of dimensional continuation, which concerns continuation
of the number of the spatial dimensions from integer to real, as well as
complex, values. From the details in this Appendix it becomes apparent
how application of this technique is hampered in the context of the
many-body theory of interacting fermions. Here we propose an alternative
scheme of dimensional continuation which, although not unique, is not
restrictive with regard to the functional dependence of the integrands
on the variable of integration.
% Sec. II:
\section{Outline of the generalised Dzyaloshinski\v{i}-Larkin scheme}
Here we present a brief outline of the generalised DL scheme. For a
detailed exposition of this scheme readers are referred to the article
by MCDC (1998). In this Section we adopt the notational conventions of
the latter work. For completeness, below $p$ and $q$ denote $d+1$-vectors
$(ip_0,{\bf p})$ and $(i q_0,{\bf q})$ respectively, with $p_0$ ($q_0$)
energy (throughout we assume $\hbar=1$) and ${\bf p}$ (${\bf q}$)
momentum in $d$-dimensional momentum space (here $d$ is integer; we
shall be explicit when $d$ is non-integer); $(ip_0,{\bf p})_{\nu}$
represents $i p_0$ when $\nu=0$ and $-p_{\nu}$ when $\nu=1,2, \dots,d$;
$\int_k f(k)$, with $f(k)\equiv F(i k_0;{\bf k})$, is the short for
$(2\pi)^{-1}\int {\rm d}k_0\,(2\pi)^{-d}\int {\rm d}^dk\; F(i k_0;{\bf k})$.
For finite temperatures $T$, $(2\pi)^{-1}\int {\rm d}k_0 (\dots)$ symbolises
a Matsubara sum, $(2\pi)^{-1}\int {\rm d}k_0\, (\dots) \to T \sum_{k_0}\;
(\dots)$, where the summation runs over the discrete fermion Matsubara
energies $2\pi i(k_0+1/2) T$, with $k_0 = 0, \pm 1,\dots$. Below, unless
specifically indicated otherwise, we assume the summation convention
$A_{\nu} B^{\nu}\equiv\sum_{\nu=0}^d A_{\nu} B^{\nu}$.
For a system whose Hamiltonian ${\hat H}$ is of the form
\begin{equation}
\label{e1}
{\hat H} \equiv {\hat H}_0 + {\hat H}_I {:=} \sum_{{\bf k}\,\sigma}
\varepsilon^0_{\bf k} {\hat a}_{{\bf k},\sigma}^{\dag}
{\hat a}_{{\bf k},\sigma}
+ \frac{1}{2 V} \sum_{{\bf k},{\bf k}',{\bf q}}\;
\sum_{\sigma\,\sigma'}\,
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\;
{\hat a}_{{\bf k}-{\bf q}/2,\sigma}^{\dag}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma}
{\hat a}_{{\bf k}'+{\bf q}/2,\sigma'}^{\dag}
{\hat a}_{{\bf k}'-{\bf q}/2,\sigma'},
\end{equation}
with ${\hat a}^{\dag}_{{\bf k},\sigma}$, ${\hat a}_{{\bf k},\sigma}$
the canonical creation and annihilation operators for fermions with
spin $\sigma$, respectively, $\varepsilon^0_{\bf k}$ the energy
dispersion of the non-interacting fermions,
\footnote{\label{f6}
We assume $\varepsilon^0_{\bf k} \equiv \|{\bf k}\|^2/2$ (in units where
$m_f$, the fermion mass, is equal to unity). For some relevant remarks
see Footnote \ref{f12} further on. }
$V$ volume of the system and $g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}
({\bf q})$ the fermion-fermion interaction function, MCDC (1998) attempt
to transform the expression for the self-energy $\Sigma_{\sigma}(p)$,
\begin{equation}
\label{e2}
\Sigma_{\sigma}(p) = -\sum_{\sigma'} \int_{p'}\;
D_{{\bf p}\,{\bf p}'}^{\sigma\,\sigma'}(p-p')
G_{\sigma}(p') \Lambda_{\sigma'\,\sigma}^0 \big([p+p']/2;p'-p),
\end{equation}
into a closed form. Here $D_{{\bf p}\,{\bf p}'}^{\sigma\,\sigma'}(p-p')$
is the screened interaction function (see Eq.~(\ref{e17}) below),
$G_{\sigma}(k) \equiv G_{\sigma}(ik_0;{\bf k})$ the (thermal)
single-particle Green function,
\begin{equation}
\label{e3}
\Lambda_{\sigma'\,\sigma}^0(p;q) {:=}
\langle {\hat n}_{\sigma'}(q) {\hat a}_{p-q/2,\sigma}
{\hat a}_{p+q/2,\sigma}^{\dag}
\rangle_{\rm tr}^{\rm irr}
\end{equation}
the irreducible density (or charge) vertex, in which
\footnote{\label{f7}
Note that $\sum_{\sigma} {\hat n}_{\sigma}(q) \equiv {\hat j}_0^0(q)
\equiv {\hat j}^0(q)$, where ${\hat j}_0^0(q)$ and ${\hat j}^0(q)$
stand for the density (or charge) parts of ${\hat j}_0^{\nu}(q)$ and
${\hat j}^{\nu}(q)$, respectively (see Eqs.~(\ref{e18}) and (\ref{e43})
below). }
\begin{equation}
\label{e4}
{\hat n}_{\sigma}({\bf q}) {:=} \sum_{\bf k}
{\hat a}_{{\bf k}-{\bf q}/2, \sigma}^{\dag}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma}.
\end{equation}
In Eq.~(\ref{e3}) ``tr'' indicates that the external Green functions
are not taken into account (i.e. they are truncated), which amounts to
dividing the expression associated with the (thermal) average $\langle
\dots\rangle^{\rm irr}$, without subscript ``tr'', by $G_{\sigma}(p+q/2)
G_{\sigma}(p-q/2)$; the superscript ``irr'' indicates that in the
diagrammatic expansion of the associated correlation function only
one-interaction-{\sl irreducible} diagrams are taken into account, i.e.
those which do not fall into two disjoint parts upon removing an
interaction line in them. Through reliance upon the generalised
closed-loop theorem (CLT) (Kopietz, Hermisson and Sch\"onhammer 1995),
\footnote{\label{f8}
The CLT (or, loop-cancellation theorem) for the one-dimensional Luttinger
(1963) model is due to Dzyaloshinski\v{i} and Larkin (1974). For an
extensive exposition of this theorem see Bohr (1981). }
MCDC (1998) relate $\Lambda_{\sigma'\,\sigma}^{\nu}(p;q)$ with the
{\sl irreducible} vertex part associated with the {\sl non-interacting}
density-current ${\hat j}_0^{\nu}(q)$, that is
\begin{equation}
\label{e5}
\Lambda_{\sigma}^{\nu}(p;q) {:=} \langle {\hat j}_0^{\nu}(q)\,
{\hat a}_{p-q/2,\sigma} {\hat a}_{p+q/2,\sigma}^{\dag}
\rangle_{\rm tr}^{\rm irr},
\end{equation}
as follows:
\begin{equation}
\label{e6}
\Lambda_{\sigma'\,\sigma}^0(p;q) = \delta_{\sigma',\sigma}\,
\Lambda_{\sigma}^0(p;q),\;\;\;\;\;
\mbox{\boldmath{$\Lambda$}}_{\sigma'\,\sigma}(p;q) =
\delta_{\sigma',\sigma}\,\Lambda_{\sigma}^0(p;q)\,
{\bf v}_{\bf p}^0.
\end{equation}
Here the $d$-dimensional vector $\mbox{\boldmath{$\Lambda$}}_{\sigma'\,
\sigma}(p;q)$ corresponds to the $\nu=1,2,\dots,d$ components of
$\Lambda_{\sigma'\,\sigma}^{\nu}(p;q)$. The expressions in Eq.~(\ref{e6})
can be written as
\begin{equation}
\label{e7}
\Lambda_{\sigma'\,\sigma}^0(p;q) = \delta_{\sigma',\sigma}\,
\frac{(iq_0,{\bf q})_{\nu}\Lambda_{\sigma}^{\nu}(p;q)}
{iq_0 - {\bf q}\cdot {\bf v}_{\bf p}^0}.
\end{equation}
The density-current vertex function $\Lambda_{\sigma}^{\nu}(p;q)$ is
related to the {\sl vertex part}
\begin{equation}
\label{e8}
\Gamma_{\sigma}^{\nu}(p;q) {:=}
\langle {\hat j}^{\nu}(q) {\hat a}_{p-q/2,\sigma}
{\hat a}^{\dag}_{p+q/2,\sigma}\rangle_{\rm tr}
\end{equation}
and the {\sl irreducible} vertex part corresponding to the {\sl
interacting} density-current ${\hat j}^{\nu}(q)$, that is
\begin{equation}
\label{e9}
\Lambda_{\sigma}^{'\,\nu}(p;q) {:=} \langle {\hat j}^{\nu}(q)\,
{\hat a}_{p-q/2,\sigma} {\hat a}_{p+q/2,\sigma}^{\dag}
\rangle_{\rm tr}^{\rm irr},
\end{equation}
through the Dyson equation
\begin{equation}
\label{e10}
\Gamma_{\sigma}^{\nu}(p;q) = \Lambda_{\sigma}^{'\,\nu}(p;q)
+ \frac{1}{V} \sum_{{\bf k}'\,\sigma'}\,
\sum_{{\bf k}''\,\sigma''}\,
J^{\nu,{\bf k}'\,\sigma'}(q)\,
g_{{\bf k}'\,{\bf k}''}^{\sigma'\,\sigma''}(q)\,
\Lambda_{\sigma}^{{\bf k}''\,\sigma''}(p;q),
\end{equation}
where
\begin{equation}
\label{e11}
J^{\nu,{\bf k}\,\sigma}(q)
{:=} -\int \frac{{\rm d}k_0}{2\pi}\;
\langle {\hat j}^{\nu}(q) {\hat a}_{k+q/2,\sigma}^{\dag}
{\hat a}_{k-q/2,\sigma} \rangle
\equiv \int \frac{{\rm d}k_0}{2\pi}\;
\Gamma_{\sigma}^{\nu}(k;q)\,
G_{\sigma}(k+q/2)\, G_{\sigma}(k-q/2),
\end{equation}
\begin{equation}
\label{e12}
\Lambda_{\sigma}^{{\bf k}\,\sigma'}(p;q) {:=}
\int \frac{{\rm d}k_0}{2\pi}\;
\langle {\hat a}_{k-q/2,\sigma'}^{\dag} {\hat a}_{k+q/2,\sigma'}
{\hat a}_{p-q/2,\sigma} {\hat a}_{p+q/2,\sigma}^{\dag}
\rangle_{\rm tr}^{\rm irr}.
\end{equation}
Owing to the {\sl exact} Ward identity
\begin{equation}
\label{e13}
(iq_0,{\bf q})_{\nu} \Gamma_{\sigma}^{\nu}(k;q)
= G_{\sigma}^{-1}(k+q/2) - G_{\sigma}^{-1}(k-q/2),
\end{equation}
from Eq.~(\ref{e11}) it readily follows that
\begin{equation}
\label{e14}
(iq_0,{\bf q})_{\nu} J^{\nu,{\bf k}\,\sigma}(q)
= {\sf n}_{\sigma}({\bf k}-{\bf q}/2)
- {\sf n}_{\sigma}({\bf k}+{\bf q}/2),
\end{equation}
where
\begin{equation}
\label{e15}
{\sf n}_{\sigma}({\bf k}\pm {\bf q}/2) {:=} \langle {\hat a}_{{\bf k}
\pm {\bf q}/2,\sigma}^{\dag} {\hat a}_{{\bf k}\pm {\bf q}/2,\sigma}
\rangle \equiv \int \frac{{\rm d}k_0}{2\pi}\; G_{\sigma}(k\pm q/2)
\end{equation}
stands for the momentum distribution function corresponding to the
spin-$\sigma$ fermions. Through relying upon (see however \S\S~III
and IV) ${\sf n}_{\sigma}({\bf k}\pm {\bf q}/2) = {\sf n}_{\sigma}
({\bf k})\pm\frac{1}{2} {\bf q}\cdot {\Bf\nabla}_{\bf k}
{\sf n}_{\sigma}({\bf k}) + {\cal O}(\|{\bf q}\|^2)$, for small
$\|{\bf q}\|$, MCDC (1998) deduce from Eq.~(\ref{e14}) ({\sl c.f.}
Eq.~(\ref{e63}) below)
\begin{equation}
\label{e16}
(i q_0, {\bf q})_{\nu} J^{\nu,{\bf k}\,\sigma}(q) = -
{\bf q}\cdot {\Bf\nabla}_{\bf k} {\sf n}_{\sigma}({\bf k})
+ {\cal O}(\|{\bf q}\|^3).
\end{equation}
The screened interaction $D_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}(q)$
in Eq.~(\ref{e2}) is that within the random-phase approximation (RPA),
\begin{equation}
\label{e17}
D_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}(q)
= g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})
+ \sum_{\sigma''} \int_{k''}\;
g_{{\bf k}\,{\bf k}''}^{\sigma\,\sigma''}({\bf q})\,
G(k''-q/2) G(k''+q/2)
D_{{\bf k}''\,{\bf k}'}^{\sigma''\,\sigma'}(q).
\end{equation}
In practice, the interacting `bubble' $G(k''-q/2) G(k''+q/2)$ is replaced
by the non-interacting one (MCDC 1998), i.e. $G_0(k''-q/2) G_0(k''+q/2)$;
this on account of the generalised CLT, under the assumption that
${\bf k}''$ is confined to a small neighbourhood of ${\bf k}_F$. For a
comment on the generalised CLT in $d > 1$ see (Farid 1999b).
From the equation of continuity $[{\hat n}({\bf q}),{\hat H}]_- = {\bf q}
\cdot {\hat {\bf j}}({\bf q})$, where
\begin{equation}
\label{e18}
{\hat n}({\bf q}) {:=}\sum_{\sigma} {\hat n}_{\sigma}({\bf q})
\end{equation}
(see Eq.~(\ref{e4}) above), one can obtain the conserved current
${\hat {\bf j}}({\bf q})$ for the system governed by the Hamiltonian
${\hat H}$ in Eq.~(\ref{e1}). The conserved current ${\hat {\bf j}}_0
({\bf q})$ corresponding to ${\hat H}_0$ in Eq.~(\ref{e1}) is obtained
through application of the continuity equation $[{\hat n}({\bf q}),
{\hat H}_0]_- = {\bf q}\cdot {\hat {\bf j}}_0({\bf q})$,
\footnote{\label{f9}
Explicit calculation yields $\big[{\hat n}({\bf q}), {\hat H}_0\big]_- =
\sum_{{\bf k}\,\sigma} \big(\varepsilon^0_{{\bf k}+{\bf q}/2}
-\varepsilon^0_{{\bf k}-{\bf q}/2}\big)\, {\hat a}^{\dag}_{{\bf k}
-{\bf q}/2,\sigma} {\hat a}_{{\bf k}+{\bf q}/2,\sigma}$. With
$\varepsilon^0_{\bf k} = \|{\bf k}\|^2/2$ (see Footnote \ref{f6} above
and Footnote \ref{f12} below), one has $\varepsilon^0_{{\bf k}
+{\bf q}/2}-\varepsilon^0_{{\bf k}-{\bf q}/2} = {\bf q}\cdot {\bf k}
\equiv {\bf q}\cdot {\bf v}^0_{\bf k}$. }
\begin{equation}
\label{e19}
{\hat {\bf j}}_0({\bf q}) = \sum_{{\bf k}\,\sigma}\, {\bf v}_{\bf k}^0\,
{\hat a}_{{\bf k}-{\bf q}/2,\sigma}^{\dag}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma},
\end{equation}
where ${\bf v}^0_{\bf k} {:=} {\Bf\nabla}_{\bf k}\varepsilon^0_{\bf k}
\equiv {\bf k}$. The commutator $[{\hat n}({\bf q}),{\hat H}_I]_-$ gives
rise to `non-diagonal' quartic contributions (involving four different
momenta) to the current operator ${\hat {\bf j}}_I({\bf q}) {:=}
{\hat {\bf j}}({\bf q}) - {\hat {\bf j}}_0({\bf q})$, which are irrelevant
in the low-energy limit (from the renormalisation-group perspective);
taking the `diagonal' contributions, corresponding to the momentum
transfer ${\bf q}'$ equal to either ${\bf q}$ or $-{\bf q}$, the
following (approximate) expression is obtained (MCDC 1998)
\begin{eqnarray}
\label{e20}
\big[{\hat n}({\bf q}),{\hat H}_I\big]_-
&=&\frac{1}{2 V} \sum_{{\bf k}\,\sigma}\,
\sum_{{\bf k}'\,\sigma'}\,
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}(-{\bf q})\,
\big[ {\hat {\sf n}}_{\sigma}({\bf k}-{\bf q}/2)
-{\hat {\sf n}}_{\sigma}({\bf k}+{\bf q}/2)\big]\,
{\hat a}_{{\bf k}'-{\bf q}/2,\sigma'}^{\dag}
{\hat a}_{{\bf k}'+{\bf q}/2,\sigma'}\nonumber\\
&+&\frac{1}{2 V} \sum_{{\bf k}\,\sigma}\,
\sum_{{\bf k}'\,\sigma'}\,
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\,
{\hat a}_{{\bf k}-{\bf q}/2,\sigma}^{\dag}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma}
\big[ {\hat {\sf n}}_{\sigma'}({\bf k}'-{\bf q}/2)
-{\hat {\sf n}}_{\sigma'}({\bf k}'+{\bf q}/2)\big].
\end{eqnarray}
MCDC (1998) cast the right-hand side (RHS) of this expression into the
form ${\bf q} \cdot {\hat {\bf j}}_I({\bf q})$, with
\begin{equation}
\label{e21}
{\hat {\bf j}}_I({\bf q}) = \sum_{{\bf k}\,\sigma}\,
{\hat {\bf v}}_{{\bf k}\,\sigma}^I({\bf q})\,
{\hat a}_{{\bf k}-{\bf q}/2,\sigma}^{\dag}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma},
\end{equation}
where
\begin{equation}
\label{e22}
{\hat {\bf v}}_{{\bf k}\,\sigma}^I({\bf q})
{:=} -\frac{1}{V} \sum_{{\bf k}'\,\sigma'}\,
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\,
{\Bf\nabla}_{{\bf k}'}\,
{\hat {\sf n}}_{\sigma'}({\bf k}').
\end{equation}
In obtaining the result in Eq.~(\ref{e21}), MCDC (1998) employ ${\hat
{\sf n}}_{\sigma}({\bf k}-{\bf q}/2) - {\hat {\sf n}}_{\sigma}({\bf k}
+{\bf q}/2) = {\bf q}\cdot [{\Bf\nabla}_{\bf k} {\hat {\sf n}}_{\sigma}
({\bf k})] + {\cal O}(\|{\bf q}\|^3)$ (see \S\S~III.B and IV however;
also compare with Eq.~(\ref{e63}) below), and the same for ${\hat
{\sf n}}_{\sigma'}({\bf k}'-{\bf q}/2)-{\hat {\sf n}}_{\sigma'}({\bf k}'
+{\bf q}/2)$. On account of the fact that for the ground and low-lying
excited states of the system $[{\Bf\nabla}_{\bf k} {\hat {\sf n}}_{\sigma}
({\bf k})]$ is most significant for ${\bf k}={\bf k}_F$ (the same
concerning $[{\Bf\nabla}_{{\bf k}'} {\hat {\sf n}}_{\sigma'}({\bf k}')]$
for ${\bf k}'={\bf k}_F'$), MCDC (1998) make use of the following
relationship concerning the radial part of the ${\bf k}'$ integration
(that is that in the direction normal to the Fermi surface) in the
expression for ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ in
Eq.~(\ref{e22}),
\footnote{\label{f10}
In the thermodynamic limit one has $V^{-1}\sum_{{\bf k}'} (\dots) \to
(2\pi)^{-d} \int {\rm d}^d{\bf k}'\; (\dots)$. For $d$-dimensional
integrations see Appendix. }
\begin{eqnarray}
\label{e23}
\int {\rm d}k_r'\; {k_r'}^{d-1}
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\,
{\Bf\nabla}_{{\bf k}'} {\hat {\sf n}}_{\sigma'}({\bf k}') &\approx&
\|{\bf k}_F'\|^{d-1} g_{{\bf k}\,{\bf k}_F'}^{\sigma\,\sigma'}({\bf q})
\int {\rm d}k_r'\; {\Bf\nabla}_{{\bf k}'}
{\hat {\sf n}}_{\sigma'}({\bf k}')\nonumber\\
&=& - \|{\bf k}_F'\|^{d-1}
g_{{\bf k}\,{\bf k}_F'}^{\sigma\,\sigma'}({\bf q})\,
{\Bf n}_{{\bf k}_F'},
\end{eqnarray}
where ${\Bf n}_{{\bf k}_F'}$ denotes the outward unit vector normal to
the Fermi surface at ${\bf k}_F'$. Here $d$ is {\sl not} necessarily
integer (see Appendix). In arriving at the last expression on the RHS of
Eq.~(\ref{e23}), MCDC (1998) have made use of the fact that for the ground
and low-lying excited states of the system, ${\hat {\sf n}}_{\sigma'}
({\bf k}')$ yields unity for ${\bf k}'$ deep inside and zero for
${\bf k}'$ far outside the Fermi sea (that is for ${\bf k}' ={\bf k}_i'$
and ${\bf k}'={\bf k}_o'$, respectively). MCDC (1998) observe that the
RHS of Eq.~(\ref{e23}) (which is a $c$-number) would remain intact if
${\hat {\sf n}}_{\sigma'}({\bf k}')$ on the left-hand side (LHS) were
the {\sl expectation value} $\langle {\hat {\sf n}}_{\sigma'}({\bf k}')
\rangle \equiv {\sf n}_{\sigma'}({\bf k}')$.
\footnote{\label{f11}
MCDC (1998) state namely ``The $k_r'$ integral over the operator
expression thus yields a $c$-number, and the same $c$-number is
obtained when replacing ${\hat {\sf n}}_{\sigma'}({\bf k}')$ by its
expectation value.''}
Thus MCDC (1998) employ ({\sl c.f.} Eq.~(\ref{e22}) above)
\begin{equation}
\label{e24}
{\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q}) \to
{\bf v}^I_{{\bf k}\,\sigma}({\bf q}) {:=}
-\frac{1}{V} \sum_{{\bf k}'\,\sigma'}
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\,
{\Bf\nabla}_{{\bf k}'} {\sf n}_{\sigma'}({\bf k}'),
\end{equation}
and consequently ({\sl c.f.} Eq.~(\ref{e21}) above)
\begin{equation}
\label{e25}
{\hat {\bf j}}_I({\bf q}) \to
{\hat {\bf j}}_I({\bf q}) {:=}
\sum_{{\bf k}\,\sigma}\,
{\bf v}_{{\bf k}\,\sigma}^I({\bf q})\,
{\hat a}_{{\bf k}-{\bf q}/2,\sigma}^{\dag}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma}.
\end{equation}
It is this ${\hat {\bf j}}_I({\bf q})$ that MCDC (1998) employ in the
construction of the generalised DL scheme. In \S~III.B we extensively
discuss the shortcomings of the foundation of this expression for
${\hat {\bf j}}_I({\bf q})$. Here we only briefly mention that,
contrary to the statement by MCDC (1998), the relationship between the
second and the third expression in Eq.~(\ref{e23}) is {\sl defective}:
momenta being confined to a narrow band of width $\lambda$ circumscribing
the Fermi surface, we have $\|{\bf k}_o - {\bf k}_i\| = \lambda$. With $0 <
\lambda \ll \|{\bf k}_F\|$, unless the ground and low-lying excited states
of the system be uncorrelated (or at most weakly correlated), it is not
permitted to replace ${\hat {\sf n}}_{\sigma'}({\bf k}_i)$ by $1$ and
${\hat {\sf n}}_{\sigma'}({\bf k}_o)$ by $0$. To clarify this point, let
us take the expectation value of the expressions in Eq.~(\ref{e23}) with
respect to the $N$-particle ground state of ${\hat H}$. With $0 <\lambda
\ll \|{\bf k}_F'\|$, it is seen that $-1$ on the RHS of Eq.~(\ref{e23})
has to be replaced by ${\sf n}_{\sigma'}({\bf k}_o) - {\sf n}_{\sigma'}
({\bf k}_i) \approx -Z_{{\bf k}_F'\,\sigma'}$, where $Z_{{\bf k}_F'\,
\sigma'}$ stands for the amount of discontinuity in ${\sf n}_{\sigma'}
({\bf k}')$ at ${\bf k}' = {\bf k}_F'$ which is less than unity for
interacting systems and is equal to zero for such NFL states as the
marginal (Varma, {\sl et al.}, 1989, 1990, Littlewood and Varma 1991)
and Luttinger liquids (Haldane 1980, 1981, Anderson 1997). This undue
replacement of $Z_{{\bf k}_F'\,\sigma'}$ by unity, is exactly the step
undertaken in the construction of the Haldane (1992) bosonisation scheme
in $d > 1$; the consequences of this replacement would not be as severe
were it not that in cases where $Z_{{\bf k}_F'\,\sigma'}$ is small, or
vanishing, the other contributions, which are neglected within the
bosonisation scheme, were not so significantly more dominant than that
of the quantum anomaly that in an enhanced way (through replacing
$Z_{{\bf k}_F'\,\sigma'}$ by unity) is taken into account (Farid 1999b).
We observe that, unless the system under consideration be non-interacting,
the constraint on $\lambda$, namely $\lambda\ll \|{\bf k}_F\|$, implies
that in contradiction to the suggestion invoked by Eq.~(\ref{e23}),
$\int {\rm d}{k'}_r\; {k'}_r^{d-1}\, g_{{\bf k}\,{\bf k}'}^{\sigma\,
\sigma'}({\bf q}) {\Bf\nabla}_{{\bf k}'} {\hat {\sf n}}_{\sigma'}
({\bf k}')$ is {\sl not} even approximately a $c$-number. Be it as it may,
as we shall demonstrate in \S~III.B, the expression in Eq.~(\ref{e21})
with ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ as defined in
Eq.~(\ref{e22}) is {\sl not} the appropriate expression for
${\hat {\bf j}}_I({\bf q})$ and, moreover, the appropriate expression
(see Eqs.~(\ref{e51}) and (\ref{e49}) below) even for large $\lambda$ {\sl
cannot} be replaced by a $c$-number.
Through applying $(iq_0,{\bf q})_{\nu}$ to both sides of Eq.~(\ref{e10})
and making use of Eqs.~(\ref{e14}), (\ref{e21}) and (\ref{e22}), MCDC
(1998) obtain the following crucial relationship
\begin{equation}
\label{e26}
(iq_0,{\bf q})_{\nu} \Gamma_{\sigma}^{\nu}(p;q)
= (iq_0,{\bf q})_{\nu} \Lambda_{\sigma}^{\nu}(p;q).
\end{equation}
In arriving at this result, use has been made of the relations
$(2\pi)^{-1}\int {\rm d}k_0'\; {\hat a}_{k'-q/2,\sigma'}^{\dag}
{\hat a}_{k'+q/2,\sigma'} = {\hat a}_{{\bf k}'-{\bf q}/2,
\sigma'}^{\dag} {\hat a}_{{\bf k}'+{\bf q}/2,\sigma'}$ and
\begin{equation}
\label{e27}
-\mbox{\boldmath{$\Lambda$}}_{\sigma}'(p;q)+\langle {\hat {\bf j}}_I(q)
{\hat a}_{p-q/2,\sigma} {\hat a}_{p+q/2,\sigma}^{\dag}
\rangle_{\rm tr}^{\rm irr}\,\equiv\,\langle {\hat {\bf j}}_0(q)
{\hat a}_{p-q/2,\sigma} {\hat a}_{p+q/2,\sigma}^{\dag}
\rangle_{\rm tr}^{\rm irr}\, {=:}\,
\mbox{\boldmath{$\Lambda$}}_{\sigma}(p;q).
\end{equation}
From the {\sl exact} Ward identity in Eq.~(\ref{e13}) one finally obtains
\begin{equation}
\label{e28}
(iq_0,{\bf q})_{\nu} \Lambda_{\sigma}^{\nu}(p;q)
= G_{\sigma}^{-1}(p+q/2) - G_{\sigma}^{-1}(p-q/2)
\end{equation}
which in combination with Eq.~(\ref{e7}) yields (MCDC 1998)
\begin{equation}
\label{e29}
\Lambda_{\sigma',\sigma}^0(p;q) = \delta_{\sigma',\sigma}\,
\frac{G_{\sigma}^{-1}(p+q/2) - G_{\sigma}^{-1}(p-q/2)}{iq_0
- {\bf q}\cdot {\bf v}_{\bf p}^0}.
\end{equation}
Eq.~(\ref{e2}) in conjunction with Eqs.~(\ref{e17}), (\ref{e29}) and
the Dyson equation leads to a closed set of equations for the
single-particle Green function $G_{\sigma}(p)$. One has
\begin{equation}
\label{e30}
G_{\sigma}(p) = G_{0;\sigma}(p)
\Big\{ 1 - \int_{p'} \frac{D_{{\bf p}\,{\bf p}'}^{\sigma\,\sigma}
(p-p')}{i (p-p_0) - {\bf v}^0_{[{\bf p}+{\bf p}']/2}\cdot
({\bf p}-{\bf p}')}\, G_{\sigma}(p')
+ X_{\sigma}(p) G_{\sigma}(p)\Big\},
\end{equation}
where $G_{0;\sigma}(p)$ denotes the non-interacting counterpart of
$G_{\sigma}(p)$ and
\begin{equation}
\label{e31}
X_{\sigma}(p) {:=}
\int_{p'} \frac{D_{{\bf p}\,{\bf p}'}^{\sigma\,\sigma}
(p-p')}{i (p-p_0) - {\bf v}^0_{[{\bf p}+{\bf p}']/2}\cdot
({\bf p}-{\bf p}')}.
\end{equation}
In practice (MCDC 1998) $D_{{\bf p}\,{\bf p}'}^{\sigma\,\sigma}(p-p')$
(see Eq.~(\ref{e17})) is approximated by $D_{{\bf p}\,{\bf p}}^{\sigma\,
\sigma} (p-p') {=:} D_{\bf p}^{\sigma\,\sigma}(p-p')$.
For the case of an isotropic system, through replacing
$\|{\bf v}^0_{\bf k}\|$ by $v_F^0$ and employing for small ${\bf q} {:=}
{\bf p} - {\bf p}'$ ({\sl c.f.} the denominators of the expressions on
the RHSs of Eqs.~(\ref{e30}) and (\ref{e31}))
\begin{equation}
\label{e32}
{\bf v}^0_{[{\bf p}+{\bf p}']/2}\cdot ({\bf p}-{\bf p}')
= v^0_F (p_r - p_r') + {\cal O}(\|{\bf q}\|^2),
\end{equation}
where $p_r {:=} \|{\bf p}\| - \|{\bf k}_F\|$, $p_r' {:=} \|{\bf p}'\|
-\|{\bf k}_F\|$, and for small $q=(i q_0, {\bf q})$,
\begin{equation}
\label{e33}
D_{\bf p}^{\sigma\,\sigma}(q) = D(q_r/q_0,q_t/q_0),
\end{equation}
where $q_r {:=} {\bf q}\cdot {\bf p}/\|{\bf p}\|$ and $q_t {:=}
(\|{\bf q}\|^2 - q_r^2)^{1/2} \geq 0$, the radial and tangential
components of ${\bf q}$, respectively, MCDC (1998) render the expression
for $G_{\sigma}(p)$ in Eq.~(\ref{e30}) suitable for treatment by the
conventional method of dimensional continuation (see Appendix --- we
note in passing that here $q_t$ is the component of ${\bf q}$ inside
the {\sl tangent space} specified in Appendix). In this way MCDC (1998)
establish that the cross-over dimension from Luttinger- to FL state is
$d=1$; that is, the metallic state of the system under consideration
is a FL (though not necessarily a conventional one) in $d > 1$. This
finding is in agreement with the $\epsilon$-expansion result by Ueda
and Rice (1984) around $d=1$; according to these authors, in the $d=1
+\epsilon$ fermion system the only {\sl weak-coupling} fixed point is
the FL one. As we shall demonstrate in \S~III.A, the expression in
Eq.~(\ref{e29}) whose denominator on the RHS is the origin of the
denominators on the RHSs of the expressions in Eqs.~(\ref{e30}) and
(\ref{e31}) (which is further simplified in Eq.~(\ref{e32})), is
{\sl not} valid except in the {\sl strictly} non-interacting case. In
\S~III.A we further indicate that for NFLs, $\Sigma_{\sigma}({\bf p}\pm
{\bf q}/2,\omega\pm\omega_0/2)$ does {\sl not} allow for an effective
disentangling of ${\bf q}$ from ${\bf p}$ (as well as $\omega_0$ from
$\omega$) and consequently for these systems the appropriate equation
for $G_{\sigma}(p)$ does {\sl not} accommodate terms involving
contributions whose form would resemble the denominator of the
integrands on the RHSs of Eqs.~(\ref{e30}) and (\ref{e31}) which thus
would be amenable to such reduction as in Eq.~(\ref{e32}). Note that
the expression in Eq.~(\ref{e32}) is specific in that it does not involve
a component of ${\bf p}'$ in the {\sl tangent subspace} $S_{\perp}$ (see
Appendix). Thus we conclude that the general formalism for systems of
interacting fermions even in the limit of strong forward scattering does
{\sl not} lend itself for treatment by the conventional method of
dimensional continuation.
In the following Section we subject the generalised DL formalism to
a critical analysis and demonstrate its failure in $d > 1$. We also
indicate the underlying reasons for the exactness of this formalism
in $d=1$ when applied to the one-dimensional Luttinger model (1963).
% Sec. III:
\section{Critical analyses}
In this Section we present our analyses of some crucial elements that
are involved in the derivation of the generalised DL method described
in \S~II.
% III.A:
\subsection{On the relationship between
\mbox{$\Lambda_{\sigma}^0(p;q)$} and
\mbox{${\bf\Lambda}_{\sigma}(p;q)$} }
Here we restrict our considerations to the case of zero temperature
($T=0$) and deal with the analytic continuations of functions of
imaginary energy along the real energy axis. We denote the analytic
continuations of such functions as $G_{\sigma}(p)$ and $\Sigma_{\sigma}
(p)$ by $G_{\sigma}({\bf p},\omega)$ and $\Sigma_{\sigma}({\bf p},\omega)$
respectively.
From the Dyson equation for the Green function it follows that
\footnote{\label{f12}
In expressions akin to that in Eq.~(\ref{e34}) (such as that presented
in Eq.~(\ref{e42}) below), MCDC (1998) add ${\cal O}(\|{\bf q}\|^2)$ to
the RHSs in order to account for the fact that in general the
non-interacting energy dispersions can be more complicated than linear
or quadratic functions of momentum. We do not follow this practice for
the reason that the one-particle kinetic-energy operator ${\hat H}_0$
and the velocity operator ${\hat {\bf v}}$ are related to one another
through the momentum operator ${\hat {\bf p}}$; we have ${\hat H}_0 {:=}
{\hat {\bf p}}^2/2$ and ${\hat {\bf v}} {:=} {\hat {\bf p}}$ (we use
$m_f=1$ throughout), so that in the momentum representation must hold
(assuming a uniform state) $\varepsilon_{\bf p}^0 \equiv \|{\bf p}\|^2/2$
and ${\bf v}^0_{\bf p} \equiv {\bf p}$ for {\sl all} ${\bf p}$ (from this
perspective $\varepsilon^0_{{\bf p}+{\bf q}/2} -\varepsilon^0_{{\bf p}
-{\bf q}/2} = {\bf q}\cdot {\bf v}^0_{\bf p}$, whence absence of ${\cal O}
(\|{\bf q}\|^2)$ in Eq.~(\ref{e42})). A fully consistent generalisation
of the formalism considered in the present work, that is one capable of
dealing with general non-interacting energy dispersions, must therefore
involve some reformulation of the problem at hand. }
\begin{eqnarray}
\label{e34}
& &G^{-1}_{\sigma}({\bf p}+{\bf q}/2,\omega+\omega_0/2)
-G^{-1}_{\sigma}({\bf p}-{\bf q}/2,\omega-\omega_0/2)
= \omega_0 - {\bf q}\cdot {\bf v}^0_{{\bf p}\,\sigma}
\nonumber\\
& &\;\;\;\;\;\;\;\;\;\;\;\;\;
-\big[\Sigma_{\sigma}({\bf p}+{\bf q}/2,\omega+\omega_0/2)
-\Sigma_{\sigma}({\bf p}-{\bf q}/2,\omega-\omega_0/2)\big].
\end{eqnarray}
{\sl Assuming} continuous differentiability of $\Sigma_{\sigma}({\bf p}',
\omega')$ with respect to ${\bf p}'$ and $\omega'$ in neighbourhoods of
${\bf p}'={\bf p}$ and $\omega'=\omega$ respectively, for sufficiently
small $\|{\bf q}\|$ and $\vert\omega_0\vert$ we have
\begin{eqnarray}
\label{e35}
G^{-1}_{\sigma}({\bf p}+{\bf q}/2,\omega+\omega_0/2) &-&
G^{-1}_{\sigma}({\bf p}-{\bf q}/2,\omega-\omega_0/2)
= \omega_0\, \big[1 -
\partial\Sigma_{\sigma}({\bf p},\omega)/\partial\omega\big]
\nonumber\\
& &\;\;
- {\bf q}\cdot \big[{\bf v}^0_{{\bf p}\,\sigma}
+ {\Bf\nabla}_{\bf p}\Sigma_{\sigma}({\bf p},\omega)\big]
+ {\cal O}\big(\|{\bf q}\|^{\gamma},\vert\omega_0\vert^{\gamma'}\big),
\end{eqnarray}
where $\gamma,\gamma' > 1$. From Eqs.~(\ref{e28}) and (\ref{e35}), up to
corrections of the form ${\cal O}\big(\|{\bf q}\|^{\gamma},\vert\omega_0
\vert^{\gamma'}\big)$, we deduce
\begin{equation}
\label{e36}
\Lambda_{\sigma}^0({\bf p},\omega) = 1 -
\frac{\partial\Sigma_{\sigma}({\bf p},\omega)}{\partial \omega},
\;\;\;\;\;
{\bf\Lambda}_{\sigma}({\bf p},\omega)
= {\bf v}^0_{{\bf p}\,\sigma}
+ {\Bf\nabla}_{\bf p}\Sigma_{\sigma}({\bf p},\omega).
\end{equation}
Following the expressions in Eq.~(\ref{e6}) relating $\Lambda_{\sigma}^0$
to ${\bf\Lambda}_{\sigma}$, the results in Eq.~(\ref{e36}) imply
\begin{equation}
\label{e37}
{\Bf\nabla}_{\bf p}\Sigma_{\sigma}({\bf p},\omega)
= -\frac{\partial\Sigma_{\sigma}({\bf p},\omega)}{\partial
\omega}\,{\bf v}^0_{{\bf p}\,\sigma},
\end{equation}
which must hold {\sl exactly} for ${\bf p}={\bf p}_F$ and $\omega=\omega_F$
if the results in Eqs.~(\ref{e7}) and (\ref{e29}) are to be asymptotically
valid for $\|{\bf q}\|\to 0$ and $\vert\omega_0\vert\to 0$. We note that a
$\Sigma_{\sigma}({\bf p},\omega)$ whose associated $\Sigma_{\sigma}
({\bf p}_F,\omega)$ and $\Sigma_{\sigma}({\bf p},\omega_F)$ are continuously
differentiable functions of $\omega$ and ${\bf p}$ in neighbourhoods of
$\omega=\omega_F$ and ${\bf p}={\bf p}_F$, respectively, by definition
corresponds to a FL (Farid 1999a). In other words, if valid, Eq.~(\ref{e37})
must hold for FLs when $\omega=\omega_F$ as ${\bf p}\to {\bf p}_F$ and
${\bf p}={\bf p}_F$ as $\omega\to\omega_F$.
Let us now consider the following asymptotic expressions which apply
to {\sl conventional} FLs in $d=2$ for ${\bf p}$ close to the Fermi
surface and $\omega\to \omega_F$ (Hodges, Smith and Wilkins 1971,
Bloom 1975)
\footnote{\label{f13}
See also Fujimoto (1990) and Fukuyama, Narikio and Hasegawa (1991).}
(in what follows we assume the system under consideration to be in the
paramagnetic phase and thus suppress the spin indices):
\begin{eqnarray}
\label{e38}
{\rm Re}\Sigma({\bf p},\omega) &\sim&
\Sigma({\bf p},\omega_F)
+ \beta_{\bf p}\, (\omega-\omega_F),\\
\label{e39}
{\rm Im}\Sigma({\bf p},\omega) &\sim&
\alpha_{\bf p}\, {\rm sgn}(\omega-\omega_F)\,
(\omega-\omega_F)^2 \ln\vert\omega-\omega_F\vert,
\end{eqnarray}
where $\beta_{\bf p}$ and $\alpha_{\bf p}$ are real valued and for
{\sl interacting} FLs $\beta_{\bf p} < 0$ (strictly negative; see
text following Eq.~(\ref{e44}) below) and $\alpha_{\bf p} \ge 0$; for
conventional FLs in $d=2$, $\Sigma({\bf p},\omega_F)\not\equiv 0$ and
$\alpha_{\bf p} > 0$. Substitution of the RHSs of Eqs.~(\ref{e38}) and
(\ref{e39}) in Eq.~(\ref{e37}) results in
\begin{eqnarray}
\label{e40}
{\Bf\nabla}_{\bf p}\Sigma({\bf p},\omega_F) +
\big[{\Bf\nabla}_{\bf p} \beta_{\bf p}\big]\,
(\omega-\omega_F) &\sim& -\beta_{\bf p}\, {\bf v}^0_{\bf p},\\
\label{e41}
\big[{\Bf\nabla}_{\bf p}\alpha_{\bf p}\big]\, (\omega-\omega_F)
\,\ln\vert\omega-\omega_F\vert &\sim&
-\big(2\ln\vert\omega-\omega_F\vert + 1\big)\,
\alpha_{\bf p}\, {\bf v}^0_{\bf p}.
\end{eqnarray}
We now consider the consequences of these results. As will become evident,
Eqs.~(\ref{e40}) and (\ref{e41}) hold only for strictly non-interacting
FLs (in this case they hold identically), thus implying incorrectness of
Eq.~(\ref{e37}) and by extension that of the expressions in Eq.~(\ref{e6})
for interacting systems. For $\omega=\omega_F$, Eq.~(\ref{e40}) leads to
${\Bf\nabla}_{\bf p} \Sigma({\bf p},\omega_F) \equiv -\beta_{\bf p}
{\bf v}^0_{\bf p}$ (for ${\bf p}$ in the vicinity of the Fermi momentum),
which, according to Eq.~(\ref{e40}) for $\omega\not=\omega_F$, implies
${\Bf\nabla}_{\bf p} \beta_{\bf p} \equiv {\bf 0} \iff \beta_{\bf p} \equiv
0$ (see further on). This result in turn necessitates ${\Bf\nabla}_{\bf p}
\Sigma({\bf p},\omega_F)\equiv {\bf 0}$; further, for $\omega\to\omega_F$
the LHS of Eq.~(\ref{e41}) vanishes while, unless $\alpha_{\bf p}\equiv 0$,
the RHS diverges. It follows that unless the system of fermions (in $d=2$)
be non-interacting, Eq.~(\ref{e37}) and, in consequence, the relations in
Eq.~(\ref{e6}) are incorrect (see text following Eq.~(\ref{e45}) below).
For non-interacting systems we have namely (see Footnote \ref{f12} above)
\begin{equation}
\label{e42}
G_0^{-1}({\bf p}+{\bf q}/2,\omega+\omega_0/2) -
G_0^{-1}({\bf p}-{\bf q}/2,\omega-\omega_0/2) \equiv
\omega_0 - {\bf q}\cdot {\bf v}^0_{\bf p},
\end{equation}
from which it is readily deduced that, with $(1,{\bf v}^0_{\bf p})^{\nu}$
the non-interacting counterpart of the irreducible vertex part
$\Lambda_{\sigma}^{\nu}(p;q)$, the relations in Eq.~(\ref{e6}) are
identically valid in the strictly non-interacting case. This completes
the proof of our statement with regard to the condition for validity of
the expressions in Eq.~(\ref{e6}). We emphasise that the case of $d=1$
is very special when ${\bf v}^0_{\bf p}$ is a constant, say ${\bf v}^0$
(which points to a definite direction), as is the case for the
one-dimensional Luttinger (1963) model. Here the relations in
Eq.~(\ref{e6}) hold identically, for whatever $\Lambda_{\sigma}^0(p;q)$
may be, from the definition for ${\hat j}_{0}^{\nu}({\bf q})$, namely
(see Eqs.~(\ref{e18}), (\ref{e4}) and (\ref{e19}) above)
\begin{equation}
\label{e43}
{\hat j}_{0}^{\nu}({\bf q}) {:=} \big({\hat n}({\bf q}),
{\hat {\bf j}}_{0}({\bf q})\big)^{\nu} \equiv
\sum_{{\bf k}\,\sigma}\, (1,{\bf v}^0_{\bf k})^{\nu}\,
{\hat a}^{\dag}_{{\bf k}-{\bf q}/2\,\sigma}
{\hat a}_{{\bf k}+{\bf q}/2\,\sigma},
\end{equation}
it follows that ${\bf\Lambda}_{\sigma}(p;q)$ is ${\bf v}^0$ times
$\Lambda_{\sigma}^0(p;q)$ (see Eq.~(\ref{e6})).
We conclude that the relations in Eq.~(\ref{e6}) are {\sl not}
asymptotically correct in the forward-scattering limit when $d > 1$,
unless the system under consideration be non-interacting, in which
case they are identically valid in {\sl any} spatial dimension.
Above we have asserted that ${\Bf\nabla}_{\bf p} \beta_{\bf p} \equiv
{\bf 0} \iff \beta_{\bf p}\equiv 0$. We now demonstrate the validity of
this equivalence. For FLs, it can be shown that (Farid 1999a)
\begin{equation}
\label{e44}
\beta_{\bf p} = \frac{1}{\pi}
\int_0^{\infty} \frac{{\rm d}\omega'}{{\omega'}^2}\;
\big[{\rm Im}\Sigma({\bf p},\omega'+\omega_F)
-{\rm Im}\Sigma({\bf p},-\omega'+\omega_F)\big].
\end{equation}
We note in passing that since for {\sl interacting} FLs the integrand
of this expression is not identically vanishing and, moreover, owing
to the assumed stability of the system ${\rm Im}\Sigma({\bf p},\omega'
+\omega_F)$ and $-{\rm Im}\Sigma({\bf p},-\omega'+\omega_F)$ are never
positive for $\omega'\geq 0$, it follows that $\beta_{\bf p} < 0$ (strictly
negative). The expression in Eq.~(\ref{e44}) can be transformed into
\footnote{\label{f14}
For FLs $\int_{\mu}^{i\infty} {\rm d}z\; (z-\mu)^{-2}\,
{\rm Im}[\Sigma({\bf p},z) -\Sigma({\bf p},2\mu -z)]$ exists, however
$\int_{\mu}^{i\infty} {\rm d}z\; (z-\mu)^{-2}\, [\Sigma({\bf p},z)
-\Sigma({\bf p},2\mu -z)]$ does not. Addition of the infinitesimal $0^+$
to $\mu$ in the lower boundary of the latter integral (see Eq.~(\ref{e45}))
formally renders this integral existent without affecting the value of
$\beta_{\bf p}$. }
\begin{equation}
\label{e45}
\beta_{\bf p} = \frac{1}{\pi}\, {\rm Im}\int_{\mu+0^+}^{i\infty}
\frac{{\rm d} z}{(z-\mu)^2}\;
\big[\Sigma({\bf p},z) - \Sigma({\bf p},2 \mu - z)\big],
\end{equation}
where $\mu\in (\mu_N,\mu_{N+1})$ with $\mu_N {:=} E_{N,0} - E_{N-1,0}
\equiv \omega_F$ and $\mu_{N+1} {:=} E_{N+1,0} - E_{N,0}$. Here $E_{M,0}$
stands for the total energy of the interacting $M$-particle ground state
and $N$ denotes the actual number of the particles in the system. The
$z$-integration in Eq.~(\ref{e45}) is carried out along an {\sl arbitrary}
contour in the upper half of the complex $z$ plane, connecting $z=\mu+0^+$
with the point of infinity. The {\sl arbitrariness} (under the mentioned
constraints) of the contour of the $z$-integration has a far-reaching
consequence: assuming differentiability of $\beta_{\bf p}$ with respect
to ${\bf p}$ (in some neighbourhood of ${\bf p}_F$), it follows from
${\Bf\nabla}_{\bf p}\beta_{\bf p} \equiv {\bf 0}$ that
${\Bf\nabla}_{\bf p}\big[\Sigma({\bf p},z)-\Sigma({\bf p},2\mu-z)\big]
\equiv {\bf 0}$ for {\sl all} $z$ in the upper half of the complex $z$
plane; by the Riemann-Schwarz reflection property with respect to the
real axis, this identity holds also for {\sl all} points in the lower
half of the complex $z$ plane.
\footnote{\label{f15}
For ${\rm Im}(z)\not=0$, $\Sigma({\bf p},z)$ satisfies the Riemann-Schwarz
reflection property $\Sigma({\bf p},z^*) \equiv \Sigma^*({\bf p},z)$
(see e.g., DuBois 1959, Luttinger 1961, Farid 1999a). }
This implies that $\Sigma({\bf p},z) \equiv \Sigma({\bf p}, 2\mu - z)$
over the entire complex plane. In view of the expressions in
Eqs.~(\ref{e38}) and (\ref{e39}), this result is seen to apply only if
$\Sigma({\bf p},z)$ is independent of $z$, a property which is specific
to {\sl non}-interacting systems. A corollary to this result is that
$\beta_{\bf p}\equiv 0$ (that is zero rather than any other constant that
${\Bf\nabla}_{\bf p}\beta_{\bf p} \equiv {\bf 0}$ would suggest in the
first instance; see the text following Eq.~(\ref{e41}) above).
\footnote{\label{f16}
Let ${\rm Re}(z) {=:}\omega > \mu$ and ${\rm Im}(z) {=:}\eta$, $\eta
\downarrow 0$. For such a $z$ we have ${\rm Im}\Sigma({\bf p}, z) \leq 0$
while ${\rm Im}\Sigma({\bf p},2\mu - z) \geq 0$ (validity of both of these
relations is necessary for the stability of the system). Satisfaction of
$\Sigma({\bf p},z) \equiv \Sigma({\bf p},2\mu -z)$, for {\sl all} $z$,
in this particular case would imply ${\rm Im}\Sigma({\bf p},\omega
+i\eta) \equiv 0$ for {\sl all} $\omega > \mu$; it similarly would
imply ${\rm Im}\Sigma({\bf p},\omega-i\eta) \equiv 0$ for {\sl all}
$\omega < \mu$. By the Kramers-Kr\"onig relation it follows that
$\Sigma({\bf p},z)$ is {\sl independent} of $z$. Note that in obtaining
this result we have {\sl not} relied upon the specific form of the
expressions in Eqs.~(\ref{e38}) and (\ref{e39}). }
Our above analysis has explicitly concerned FLs (through our explicit
use of the relations in Eqs.~(\ref{e38}) and (\ref{e39}), but more
importantly Eq.~(\ref{e44})), however, as can be easily verified, for
NFLs the RHS of Eq.~(\ref{e34}) does not allow for any effective
separation of ${\bf q}$ and $\omega_0$ from ${\bf p}$ and $\omega$
respectively, so that in $d > 1$ the relations in Eq.~(\ref{e6}) lack
even the most fundamental basis for validity as regards NFLs.
% III.B:
\subsection{On the relationship between
\mbox{$\Lambda_{\sigma}^{\nu}(p;q)$} and
\mbox{${\Gamma}_{\sigma}^{\nu}(p;q)$} }
Evidently for obtaining the ${\hat {\bf j}}_I({\bf q})$ as presented in
Eq.~(\ref{e21}) in terms of ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$
in Eq.~(\ref{e22}), the functions in the square brackets on the RHS of
Eq.~(\ref{e20}) have been expanded to linear order in ${\bf q}$. Such an
expansion is {\sl in general} invalid for reasons that we discuss in some
detail in \S~IV. A second step involved in the derivation of ${\hat {\bf
j}}_I({\bf q})$ in Eq.~(\ref{e25}) is that of replacing ${\Bf\nabla}_{\bf k}
{\hat {\sf n}}_{\sigma}({\bf k})$ by the expectation value
${\Bf\nabla}_{\bf k} \langle {\hat {\sf n}}_{\sigma}({\bf k})\rangle$.
The justification that has been given by MCDC (1998) for this substitution
is untenable (in particular in $d > 1$) (see also text following
Eq.~(\ref{e25}) above). For obtaining a generally valid expression for
${\hat {\bf j}}_I({\bf q})$, the RHS of Eq.~(\ref{e20}) must be first
written in terms of four separate contributions, each involving only one
of ${\hat {\sf n}}_{\sigma}({\bf k}-{\bf q}/2)$, ${\hat {\sf n}}_{\sigma}
({\bf k}+{\bf q}/2)$, ${\hat {\sf n}}_{\sigma'}({\bf k}'-{\bf q}/2)$ and
${\hat {\sf n}}_{\sigma'}({\bf k}'+{\bf q}/2)$. Subsequently, in each of
these contributions, by appropriate transformations of the summation
variables ${\bf k}$ and ${\bf k}'$, the vector ${\bf q}$ in the arguments
of these functions must be transferred into that of the interaction function
$g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}(\mp {\bf q})$, which we assume
to be a continuously differentiable function of its indices ${\bf k}$
and ${\bf k}'$ ({\sl c.f.} statement preceding Eq.~(7) in Castellani, Di
Castro and Metzner 1994). In this way one obtains (for small $\|{\bf q}\|$)
\begin{eqnarray}
\label{e46}
{\hat {\bf j}}_I({\bf q}) &=& \frac{1}{2 V} \sum_{{\bf k}\,\sigma}\,
\sum_{{\bf k}'\,\sigma'}\,\Big\{
\big[{\Bf\nabla}_{{\bf k}}
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}(-{\bf q})\big]\,
{\hat {\sf n}}_{\sigma}({\bf k})\,
{\hat a}_{{\bf k}'-{\bf q}/2,\sigma'}^{\dag}
{\hat a}_{{\bf k}'+{\bf q}/2,\sigma'}
\nonumber\\
& & \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;
+\big[{\Bf\nabla}_{{\bf k}'}
g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\big]\,
{\hat a}_{{\bf k}-{\bf q}/2,\sigma}^{\dag}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma}\,
{\hat {\sf n}}_{\sigma'}({\bf k}')\Big\}.
\end{eqnarray}
Making use of the commutation relation
\begin{equation}
\label{e47}
\Big[{\hat a}^{\dag}_{{\bf k}-{\bf q}/2,\sigma}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma},
{\hat {\sf n}}_{\sigma'}({\bf k}')\Big]_-
= \delta_{\sigma,\sigma'}\,
\big\{\delta_{{\bf k}+{\bf q}/2,{\bf k}'}
- \delta_{{\bf k}-{\bf q}/2,{\bf k}'}\big\}\,
{\hat a}^{\dag}_{{\bf k}-{\bf q}/2,\sigma}\,
{\hat a}_{{\bf k}+{\bf q}/2,\sigma},
\end{equation}
which is obtained through a repeated application of the canonical
anti-commutation relations for ${\hat a}_{{\bf k},\sigma}^{\dag}$ and
${\hat a}_{{\bf k},\sigma}$, the expression in Eq.~(\ref{e46}) can
be written as
\begin{eqnarray}
\label{e48}
{\hat {\bf j}}_I({\bf q}) &=& \sum_{{\bf k}\,\sigma}\,
{\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})\,
{\hat a}^{\dag}_{{\bf k}-{\bf q}/2,\sigma}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma} \nonumber\\
& &+\frac{1}{2 V} \sum_{{\bf k}\,\sigma}
\Big\{ \big[{\Bf\nabla}_{{\bf k}'}
\left. g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma}({\bf q})\right|_{{\bf k}'=
{\bf k}+{\bf q}/2}\big]
- \big[{\Bf\nabla}_{{\bf k}'}
\left. g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma}({\bf q})\right|_{{\bf k}'=
{\bf k}-{\bf q}/2}\big]\Big\}\,
{\hat a}^{\dag}_{{\bf k}-{\bf q}/2,\sigma}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma},
\end{eqnarray}
where ({\sl c.f.} Eq.~(\ref{e22}) above)
\begin{equation}
\label{e49}
{\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})
{:=} \frac{1}{V} \sum_{{\bf k}'\,\sigma'}
\big[ {\Bf\nabla}_{{\bf k}'}
{\bar g}_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}
({\bf q})\big]\, {\hat {\sf n}}_{\sigma'}({\bf k}'),
\end{equation}
in which
\begin{equation}
\label{e50}
{\bar g}_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q}) {:=}
\frac{1}{2} \Big( g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q}) +
g_{{\bf k}'\,{\bf k}}^{\sigma'\,\sigma}(-{\bf q})\Big).
\end{equation}
By assuming ${\Bf\nabla}_{{\bf k}'} g_{{\bf k}\,{\bf k}'}^{\sigma,
\sigma}({\bf q})$ to be a continuously-differentiable function of
${\bf k}'$, the last term on the RHS of Eq.~(\ref{e48}) is seen to be
proportional to ${\bf q}$, so that in the limit of ${\bf q}\to {\bf 0}$
it can be neglected in comparison with the first term. For small
$\|{\bf q}\|$ we can thus write
\begin{equation}
\label{e51}
{\hat {\bf j}}_I({\bf q}) = \sum_{{\bf k}\,\sigma}\,
{\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})\,
{\hat a}^{\dag}_{{\bf k}-{\bf q}/2,\sigma}
{\hat a}_{{\bf k}+{\bf q}/2,\sigma},
\end{equation}
where ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ is defined in
Eq.~(\ref{e49}). Note in passing, that for a constant interaction (i.e. a
${\bf k},{\bf k}'$-independent $g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}
({\bf q})$), ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q}) \equiv {\hat
{\bf 0}}$, so that Eq.~(\ref{e51}) correctly yields ${\hat {\bf j}}_I
({\bf q}) \equiv {\hat {\bf 0}}$, or ${\hat {\bf j}}({\bf q}) \equiv
{\hat {\bf j}}_0({\bf q})$.
The correlated velocity operator in Eq.~(\ref{e49}) is distinct from
that in Eq.~(\ref{e22}). To bring out the fundamental difference between
the two, let us follow MCDC (1998) and replace the operator ${\hat
{\sf n}}_{\sigma'}({\bf k}')$ on the RHS of Eq.~(\ref{e49}) by its
expectation value ${\sf n}_{\sigma'}({\bf k}')$ and compare the resulting
vector ${\bf v}^I_{{\bf k}\,\sigma}({\bf q})$ with that in Eq.~(\ref{e24}).
Since by assumption ${\Bf\nabla}_{{\bf k}'} {\bar g}_{{\bf k}\,
{\bf k}'}^{\sigma\,\sigma'}({\bf q})$ is a smooth function of ${\bf k}'$
(see text preceding Eq.~(\ref{e46}) above), we can write (recall that
$\|{\bf k}_o' - {\bf k}_i'\| = \lambda \ll \|{\bf k}_F'\|$)
\begin{equation}
\label{e52}
\int {\rm d}k_r'\; {k_r'}^{d-1}\,
\big[{\Bf\nabla}_{{\bf k}'} {\bar g}_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}
({\bf q})\big] {\sf n}_{\sigma'}({\bf k}')
\approx \big[\left. {\Bf\nabla}_{{\bf k}'}
{\bar g}_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\right|_{{\bf k}'
= {\bf k}_F'}\big] \int {\rm d}k_r'\; {k_r'}^{d-1}\,
{\sf n}_{\sigma'}({\bf k}').
\end{equation}
For the non-interacting case, where ${\sf n}_{\sigma'}({\bf k}') =
\Theta({\bf k}_F'- {\bf k}')$ (recall that distances are measured along
${\Bf n}_{{\bf k}_F'}$), the integral on the RHS of Eq.~(\ref{e52}) is
equal to $\big(\|{\bf k}_F'\|^d - \|{\bf k}_i'\|^d \big)/d$, so that the
RHS of Eq.~(\ref{e52}) becomes equal to $\big[{\Bf\nabla}_{{\bf k}'}
{\bar g}_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})\vert_{{\bf k}'
={\bf k}_F'}\big]\, \big(\|{\bf k}_F'\|^d - \|{\bf k}_i'\|^d\big)/d$, to
be compared with the RHS of Eq.~(\ref{e23}), which is equally applicable
to the non-interacting ground states, namely $-g_{{\bf k}\,
{\bf k}_F'}^{\sigma\,\sigma'}({\bf q})\,\|{\bf k}_F'\|^d\,
{\Bf n}_{{\bf k}_F'}$. The two results are manifestly different. This
demonstrates that Eq.~(\ref{e24}), and therefore Eq.~(\ref{e25}), is
fundamentally incorrect. This aspect is {\sl additional} to the
shortcoming of the result in Eq.~(\ref{e23}), which as we have discussed
in the text following Eq.~(\ref{e25}), shows up when the ground and
low-lying excited states of the system are (strongly) correlated. Thus
we have unequivocally established that not only the change in
Eq.~(\ref{e25}), corresponding to replacing ${\hat {\bf v}}^I_{{\bf k}
\,\sigma}({\bf q})$ as defined in Eq.~(\ref{e22}) by ${\bf v}^I_{{\bf k}
\,\sigma}({\bf q})$ as defined in Eq.~(\ref{e24}), lacks justification
(unless the system under consideration be non-interacting), but also
that the expression for ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$
as presented in Eq.~(\ref{e22}) (to be compared with the
${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ in Eq.~(\ref{e49})) is
{\sl invalid}, this on account of the fact that it relies on the Taylor
expansion of an operator (that is ${\hat {\sf n}}_{\sigma}({\bf k}\pm
{\bf q}/2)$) whose, in particular, expectation values with respect to
ground states are in general non-differentiable (see \S~IV).
A major consequence of use of Eq.~(\ref{e51}), or Eq.~(\ref{e21}),
rather than one in which the {\sl operator} ${\hat {\bf v}}^I_{{\bf k}\,
\sigma}({\bf q})$ is replaced by the $c$-number ${\bf v}^I_{{\bf k}\,
\sigma}({\bf q})$ (effected through substituting ${\sf n}_{\sigma'}
({\bf k}')$ for ${\hat {\sf n}}_{\sigma'}({\bf k}')$ in Eq.~(\ref{e49}),
or Eq.~(\ref{e22})), is violation of Eq.~(\ref{e26}) which is the crucial
link between the {\sl exact} Ward identity in Eq.~(\ref{e13}) and the
expression in Eq.~(\ref{e28}). To make this statement explicit, we first
apply $(i q_o,{\bf q})_{\nu}$ to both sides of Eq.~(\ref{e10}) and
subsequently substitute the expression on the RHS of Eq.~(\ref{e14}) for
$(i q_0,{\bf q})_{\nu} J^{\nu,{\bf k}'\,\sigma'}(q)$. Note that in doing
so, we depart from the approach by MCDC (1998) who employ the expression
on the RHS of Eq.~(\ref{e16}). Next, we write the expression for $(i q_0,
{\bf q})_{\nu} J^{\nu,{\bf k}'\,\sigma'}(q)$ in the contribution
corresponding to $(i q_0,{\bf q})_{\nu} \Gamma_{\sigma}^{\nu}(p;q)$ as
two separate terms and in each of these by transformations of the dummy
variables of summation, remove ${\bf q}/2$ from arguments of
${\sf n}_{\sigma'}({\bf k}'-{\bf q}/2)$ and ${\sf n}_{\sigma'}({\bf k}'
+{\bf q}/2)$. This procedure is exactly the same as that which we
employed for deducing ${\hat {\bf j}}_I({\bf q})$ from the equation of
continuity associated with ${\hat H}_I$ (see Eq.~(\ref{e46}) above and
text following it). Thus we obtain
\begin{eqnarray}
\label{e53}
& &\frac{1}{V} \sum_{{\bf k}'\,\sigma'} \sum_{{\bf k}''\,\sigma''}
(i q_0, {\bf q})_{\nu} J^{\nu,{\bf k}'\,\sigma'}(q)\,
g_{{\bf k}'\,{\bf k}''}^{\sigma'\,\sigma''}(q)\,
\Lambda_{\sigma}^{{\bf k}''\,\sigma''}(p;q)\nonumber\\
& &\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;
= {\bf q}\cdot \sum_{{\bf k}''\,\sigma''}\,\Big\{ \frac{1}{V}
\sum_{{\bf k}'\,\sigma'}
{\sf n}_{\sigma'}({\bf k}')\,
\big[{\Bf\nabla}_{{\bf k}'}
g_{{\bf k}'\,{\bf k}''}^{\sigma'\,\sigma''}({\bf q})
\big]\Big\}\,\Lambda_{\sigma}^{{\bf k}''\,\sigma''}(p;q).
\end{eqnarray}
Neglecting the {\sl possible} difference between $g_{{\bf k}\,
{\bf k}'}^{\sigma\,\sigma'}({\bf q})$ and ${\bar g}_{{\bf k}\,
{\bf k}'}^{\sigma\,\sigma'}({\bf q})$ (see Eq.~(\ref{e50}) above), it is
seen that the term enclosed by curly braces in Eq.~(\ref{e53}) is equal to
the expectation value of ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ as
presented in Eq.~(\ref{e49}). As we have extensively discussed above, there
exists {\sl no} justification, whatever, for ${\bf v}^I_{{\bf k}\,\sigma}
({\bf q})$ to be a good representation of ${\hat {\bf v}}^I_{{\bf k}\,
\sigma}({\bf q})$. If, as in the treatment by MCDC (1998),
${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ could be replaced by its
expectation value, then making use of the defining expression for
$\Lambda_{\sigma}^{{\bf k}''\,\sigma''}(p;q)$ in Eq.~(\ref{e12}) and the
expression for ${\hat {\bf j}}_I({\bf q})$ in Eq.~(\ref{e51}) wherein
${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ is replaced by
${\bf v}^I_{{\bf k}\,\sigma}({\bf q})$, the RHS of Eq.~(\ref{e53}) would
reduce to ${\bf q} \cdot {\Bf\Lambda}_{\sigma}(p;q)$. Adding this result
to $(i q_0,{\bf q})_{\nu} {{\Bf\Lambda}_{\sigma}'}^{\nu}(p;q)$ (see
Eq.~(\ref{e10}), to whose both sides $(i q_0,{\bf q})_{\nu}$ is applied),
one would arrive at Eq.~(\ref{e26}). We emphasise that in arriving at
Eq.~(\ref{e26}) it is the substitution of ${\hat {\bf v}}^I_{{\bf k}\,\sigma}
({\bf q})$ by ${\bf v}^I_{{\bf k}\,\sigma}({\bf q})$ that plays the vital
role and {\sl not} whether ${\bf v}^I_{{\bf k}\,\sigma} ({\bf q})$ is the
expectation value of ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ as
defined in Eq.~(\ref{e22}) or Eq.~(\ref{e49}). However, it should be
realised that contrary to the case of the ${\hat {\bf v}}^I_{{\bf k}\,
\sigma}({\bf q})$ in Eq.~(\ref{e22}) whose replacement by its expectation
value can be justified for non-interacting systems (and by extension,
weakly interacting systems) (see \S~II, text following Eq.~(\ref{e25})),
replacement of the ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ in
Eq.~(\ref{e49}), which we have demonstrated to be the correct velocity
operator, by its expectation value is {\sl not} justified, not even for
cases where the ground and low-lying excited states of the system are
non-interacting (or weakly interacting). Of course, in the non-interacting
case $g_{{\bf k}\,{\bf k}'}^{\sigma\,\sigma'}({\bf q})$ is identically
vanishing so that the problem associated with the inadmissibility of
replacing ${\hat {\bf v}}^I_{{\bf k}\,\sigma}({\bf q})$ by
${\bf v}^I_{{\bf k}\,\sigma}({\bf q})$ does not arise in this limit
(see Eq.~(\ref{e49}) above).
Above we have demonstrated that in $d > 1$, the trivial relationship
between the vertex parts $\Lambda_{\sigma}^{\nu}(p;q)$ and
$\Gamma_{\sigma}^{\nu}(p;q)$ as given in Eq.~(\ref{e26}) is invalid and
consequently {\sl no} direct link exists between the {\sl exact} Ward
identity in Eq.~(\ref{e13}) and the expression in Eq.~(\ref{e28}). Further,
our considerations in \S~III.A made evident that also Eq.~(\ref{e7}) has
no rightful place in the theory of interacting fermions in $d >1$.
% Sec IV:
\section{Mathematical transgressions and ``blessings'' of incorrect
error estimation}
In this Section we illustrate how in $d=1$, specifically in dealing
with the one-dimensional Luttinger (1963) model at $T=0$, a combination
of transgressions against mathematical rules and uncritical neglect of
truncation errors can give rise to outcomes which coincide with the
exact results, thus creating the false impression that this approach
were justified. The particular example that we consider in this Section
makes explicit that specifically in $d > 1$ the same procedure yields
manifestly incorrect results. Our illustrative example involves the
{\sl ground-state} momentum distribution function ${\sf n}({\bf k})$
which is singular at ${\bf k}={\bf k}_F$ both in the case of FLs and NFLs.
We first explicitly deal with the one-dimensional Luttinger model for
spin-less fermions (Luttinger 1963, Mattis and Lieb 1965), characterised
by the anomalous dimension $\alpha$, $\alpha \in (0,1)$. We note in
passing that for the one-dimensional Hubbard model $0 < \alpha < 1/8$,
with $\alpha=1/8$ corresponding to the limit of infinite on-site
repulsion (Voit 1993). By choosing the positive direction to be from
left to right, below we consider the one-dimensional {\sl vectors}
${\bf k}$ and ${\bf q}$ as {\sl scalars} and $k_F\equiv \vert
{\bf k}_F\vert$. Here the ground-state momentum-distribution function
${\sf n}_r({\bf k})$, with $r=\pm$ the branch index ($r=+$ indicates
the branch of the right movers and $r=-$ that of the left movers), is
continuous but {\sl not} differentiable at ${\bf k}= r k_F$. For the
density-density correlation function $\Pi_{r\,r}(q)$ pertaining to the
{\sl interacting} system we have (DL 1974; see, e.g., Metzner and
Di Castro 1993, Voit 1994)
\footnote{\label{f17}
\small
The boundaries of the integral $g_r({\bf q})$ are determined by the
observation that ${\bf k}$ in ${\sf n}_r({\bf k})$ is restricted to
the range $-\lambda/2 + r k_F \leq {\bf k} \leq \lambda/2 + r k_F$.}
\begin{equation}
\label{e54}
\Pi_{r\,r}(q) = \frac{1}{2\pi}\,
\frac{g_r({\bf q})}{i q_0- {\bf v}_r^0\cdot
{\bf q}}, \;\;\;\;\;\;
g_{r}({\bf q}) {:=} \int_{-\lambda/2 + r k_F +{\bf q}/2}^{\lambda/2
+ r k_F -{\bf q}/2} {\rm d}{\bf k}\;
\big[{\sf n}_{r}({\bf k}-{\bf q}/2) -
{\sf n}_{r}({\bf k}+{\bf q}/2)\big],
\end{equation}
where ${\bf v}_r^0$ stands for the constant velocity of the
non-interacting fermions; it points to the positive direction for
$r=+$ and the negative direction for $r=-$. Further, $\lambda$ is a
cut-off which is finite but can be arbitrarily large; for our following
considerations it is required that $\lambda$ be sufficiently large so
that ${\sf n}_r({\bf k})$ can be replaced by zero (unity) for $r=+$
($r=-$) when ${\bf k} \approx \lambda/2$. The freedom to chose $\lambda$
arbitrarily large is a specific aspect of the one-dimensional Luttinger
model. We note that although introduction of such a cut-off is necessary
for a correct treatment of the Luttinger model (Mattis and Lieb 1965),
imposition of a `sharp' cut-off on ${\sf n}_r({\bf k})$, implying
${\sf n}_{\pm}(\mp\lambda/2\mp 0^+) = 0$ and ${\sf n}_{\pm}(\mp\lambda/2
\pm 0^+) = 1$, gives rise to an incorrect asymptotic result for the
long-distance behaviour of in particular the single-particle Green
functions $G_r$ pertaining to the model under consideration;
\footnote{\label{f18}
By imposing a `sharp' cut-off on ${\sf n}_r({\bf k})$, the long-distance
behaviour of $G_r$ coincides with that specific to FLs (Farid 1999b);
such a behaviour is incorrect as long as $\alpha\not=0$. }
for this long-distance asymptotic behaviour to be described correctly,
it is necessary that ${\sf n}_+(-{\bf k})$ and ${\sf n}_-({\bf k})$ are
reduced to zero for ${\bf k}\gg \lambda/2$, which can be achieved through
imposing a `soft' cut-off on these functions which becomes operative for
$\vert {\bf k}\vert$ larger than $\lambda/2$ (Farid 1999b).
Through decomposing the integral in Eq.~(\ref{e54}) into two integrals
involving separately ${\sf n}_{r}({\bf k}-{\bf q}/2)$ and ${\sf n}_{r}
({\bf k}+{\bf q}/2)$ and subsequently transforming the integration variable
${\bf k}$, with ${\bf k}\rightharpoonup {\bf k}+{\bf q}/2$ in the first
and ${\bf k}\rightharpoonup {\bf k}-{\bf q}/2$ in the second integral,
cancellation of a common contribution to the resulting integrals gives rise
to $g_r({\bf q}) = \int_{-\lambda/2 +r k_F}^{-\lambda/2 +r k_F +{\bf q}}
{\rm d}{\bf k}\; {\sf n}_{r}({\bf k}) - \int_{\lambda/2 + r k_F
-{\bf q}}^{\lambda/2 + r k_F} {\rm d}{\bf k}\; {\sf n}_{r}({\bf k})$.
By assuming $\lambda$ to be sufficiently large (see above), ${\sf n}_{r}
({\bf k})$ in the first integral can be replaced by unity (zero) and in
the second integral by zero (unity) for $r=+$ ($r=-$), upon which one
readily obtains
\begin{equation}
\label{e55}
g_r({\bf q}) = r {\bf q}.
\end{equation}
From this and Eq.~(\ref{e54}) it follows that
\begin{equation}
\label{e56}
\Pi_{r\,r}(q) = \frac{1}{2\pi}\,
\frac{r {\bf q}}{i q_0 - {\bf v}_r^0\cdot {\bf q}}
\equiv \Pi_{r\,r}^0(q),
\end{equation}
where $\Pi_{r\,r}^0(q)$ stands for the density-density correlation
function pertaining to the {\sl non}-interacting system. This result is
a direct manifestation of the CLT (DL 1974, Bohr 1981) which applies to
the one-dimensional Luttinger model. Note that validity of Eq.~(\ref{e55})
does {\sl not} depend on any other aspect of ${\sf n}_r({\bf k})$ than it
be zero (unity) for $r=+$ ($r=-$) when ${\bf k}$ is sufficiently far to the
{\sl right} of $r k_F$ and unity (zero) for $r=+$ ($r=-$) when ${\bf k}$
is sufficiently far to the {\sl left} of $r k_F$. Thus Eq.~(\ref{e55})
holds also for the case where ${\sf n}_r({\bf k})$ in Eq.~(\ref{e54}) is
replaced by its non-interacting counter-part $\Theta(k_F-r {\bf k})$;
using this step function, the validity of this statement is trivially
verified by explicit calculation.
Let us now assume $\vert {\bf q}\vert$ to be small. Using (inappropriately)
${\sf n}_{r}({\bf k}-{\bf q}/2) - {\sf n}_{r}({\bf k}+{\bf q}/2)= -{\bf q}
\,\partial {\sf n}_{r}({\bf k})/\partial {\bf k} + {\cal O}(\vert {\bf q}
\vert^3)$ for {\sl all} ${\bf k}$ ({\sl c.f.} Eq.~(\ref{e16}) above and
Eq.~(\ref{e63}) below), through integration-by-parts and making use of
${\sf n}_{r}(\lambda/2 + r k_F -{\bf q}/2) = 0\, (1)$ and ${\sf n}_{r}
(-\lambda/2 + r k_F +{\bf q}/2)=1\, (0)$ for $r=+$ ($r=-$), up to
${\cal O}(\vert {\bf q}\vert^3)$ we evidently obtain the (exact) result
in Eq.~(\ref{e55}), in spite of the fact that ${\sf n}_{r}({\bf k})$ is
{\sl not} a differentiable function of ${\bf k}$ in a neighbourhood of
${\bf k}=r k_F$.
We shall now demonstrate that ${\cal O}(\vert {\bf q}\vert^3)$ in
$g_{r}({\bf q}) = r {\bf q} + {\cal O}(\vert {\bf q}\vert^3)$ is
incorrect. The remainder in fact is of the form ${\cal O}(\vert
{\bf q}\vert^{1+\alpha})$. Further, as the non-analytic form of this
remainder suggests, the second- and higher-order expansions of
${\sf n}_{r}({\bf k}-{\bf q}/2) - {\sf n}_{r}({\bf k}+{\bf q}/2)$
(in powers of ${\bf q}$) result in divergent integrals, reflecting
the fact that $1 < 1+\alpha < 2$.
Below we consider $\int_{-\lambda_0/2 + r k_F + {\bf q}/2}^{\lambda_0/2
+ r k_F - {\bf q}/2} {\rm d}{\bf k}\; \big[ {\sf n}_r({\bf k} -
{\bf q}/2) - {\sf n}_r({\bf k} + {\bf q}/2)\big] \equiv \int_{-\lambda_0/2
+ {\bf q}/2}^{\lambda_0/2 - {\bf q}/2} {\rm d}{\bf k}\; \big[
{\sf n}_r({\bf k} + r k_F - {\bf q}/2) - {\sf n}_r({\bf k} + r k_F +
{\bf q}/2)\big]$ in terms of the following three constituent parts
\begin{eqnarray}
\label{e57}
{\cal I}_1 &{:=}& \int_{-\lambda_0/2 + {\bf q}/2}^{-{\bf q}/2}
{\rm d}{\bf k}\; \big[ {\sf n}_r({\bf k} + r k_F - {\bf q}/2)
- {\sf n}_r({\bf k} + r k_F + {\bf q}/2)\big],\\
\label{e58}
{\cal I}_2 &{:=}& \int_{-{\bf q}/2}^{{\bf q}/2}
{\rm d}{\bf k}\; \big[ {\sf n}_r({\bf k} + r k_F - {\bf q}/2)
- {\sf n}_r({\bf k} + r k_F + {\bf q}/2)\big],\\
\label{e59}
{\cal I}_3 &{:=}& \int_{{\bf q}/2}^{\lambda_0/2 -{\bf q}/2}
{\rm d}{\bf k}\; \big[ {\sf n}_r({\bf k} + r k_F - {\bf q}/2)
- {\sf n}_r({\bf k} + r k_F + {\bf q}/2)\big],
\end{eqnarray}
where $\lambda_0$ is a constant which we chose to satisfy $\vert {\bf q}
\vert < \lambda_0 \ll \lambda$. Our choice of the inequality $\lambda_0
\ll\lambda$ is motivated by the fact that it justifies our use of the
following two-term asymptotic expression for ${\sf n}_r({\bf k})$
pertaining to the one-dimensional Luttinger model for spin-less
fermions (Voit 1994)
\begin{equation}
\label{e60}
{\sf n}_r({\bf k}+ r k_F) \sim \frac{1}{2} -
r\, C\,{\rm sgn}({\bf k})\,\vert {\bf k}\vert^{\alpha},
\end{equation}
where $C$ is a positive constant. Making use of the expression in
Eq.~(\ref{e60}), through {\sl exact} evaluation of the integrals one
obtains (below we assume ${\bf q} > 0$):
\begin{eqnarray}
\label{e61}
{\cal I}_1 = {\cal I}_3 &\sim&
\left\{ \begin{array}{ll}
\frac{r C}{\alpha+1}
\Big( (\lambda_0/2)^{\alpha+1} - (\lambda_0/2 - {\bf q})^{\alpha+1}
- {\bf q}^{\alpha+1}\Big),
&\;\;\;\;\mbox{\rm (No expansion)}\\
r\, C\, \Big( (\lambda_0/2 - {\bf q}/2)^{\alpha} -
({\bf q}/2)^{\alpha}\Big)\, {\bf q},
&\;\;\;\;\mbox{\rm (Expansion)}
\end{array}\right.\\
\label{e62}
{\cal I}_2 &\sim&
\left\{ \begin{array}{ll}
\frac{2 r C}{\alpha+1} {\bf q}^{\alpha+1},
&\;\;\;\;\mbox{\rm (No expansion)}\\
\frac{2^{1-\alpha} (\alpha + 2) r C}{\alpha +1}\,
{\bf q}^{\alpha+1}.
&\;\;\;\;\mbox{\rm (Expansion)}
\end{array}\right.
\end{eqnarray}
The dissimilarity of the results based upon the first-order expansion of
${\sf n}_r({\bf k} + r k_F - {\bf q}/2) - {\sf n}_r({\bf k} + r k_F +
{\bf q}/2)$, indicated by ``Expansion'', from the exact results, indicated
by ``No expansion'', confirms our above statement concerning the invalidity
of the expansion procedure. Note for instance that ${\cal I}_2$ as
determined through expansion coincides with the exact result only when
$\alpha=2$. It is interesting also to note that contrary to the expression
${\sf n}_{r}({\bf k}-{\bf q}/2) - {\sf n}_{r}({\bf k}+{\bf q}/2) =
-{\bf q}\,\partial {\sf n}_{r}({\bf k})/\partial {\bf k} + {\cal O}(\vert
{\bf q}\vert^3)$ which implies absence of a zeroth-order term in ${\bf q}$,
from Eq.~(\ref{e60}) one in fact has ({\sl c.f.} Eq.~(\ref{e16}) above)
\begin{eqnarray}
\label{e63}
{\sf n}_{r}({\bf k}-{\bf q}/2)-{\sf n}_{r}({\bf k}+{\bf q}/2)
&\approx& 2 r C\,{\rm sgn}({\bf q})\,
\vert {\bf k} - r k_F \vert^{\alpha} +
\alpha r C \vert {\bf k}-r k_F\vert^{\alpha-1}\, {\bf q},\nonumber\\
& &\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;
\mbox{\rm when}\;\;
{\bf k} \in [r k_F -{\bf q}/2, r k_F+{\bf q}/2],
\end{eqnarray}
which contains a zeroth-order term in ${\bf q}$ as long as ${\bf k}
\not= r k_F$. Without taking into account the zeroth-order term in
Eq.~(\ref{e63}), we would have ${\cal I}_2 \sim 2^{1-\alpha} r C
{\bf q}^{\alpha+1}$ (Expansion) rather than the `(Expansion)' result for
${\cal I}_2$ in Eq.~(\ref{e62}); the two results coincide only for
$\alpha=1$.
Adding the three contributions ${\cal I}_j$, $j=1,2,3$, as presented
in Eqs.~(\ref{e61}) and (\ref{e62}), we obtain
\begin{eqnarray}
\label{e64}
& &\int_{-\lambda_0/2 + r k_F + {\bf q}/2}^{\lambda_0/2 + r k_F
- {\bf q}/2} {\rm d}{\bf k}\;
\big[ {\sf n}_r({\bf k}-{\bf q}/2) -
{\sf n}_r({\bf k}+{\bf q}/2)\big]\nonumber\\
& &\;\;\;\;\;\;\;\;\;\;\;\;
\sim \left\{ \begin{array}{ll}
\frac{2 r C}{\alpha+1}
\Big( (\lambda_0/2)^{\alpha+1}
- (\lambda_0/2 - {\bf q})^{\alpha+1}\Big),
&\;\;\;\mbox{\rm (No expansion)}\\
\frac{2^{1-\alpha} (\alpha+2) r C}{\alpha+1}\,
{\bf q}^{\alpha+1} + 2 r C \Big( (\lambda_0/2 -
{\bf q}/2)^{\alpha} - ({\bf q}/2)^{\alpha} \Big).
&\;\;\;\mbox{\rm (Expansion)}
\end{array} \right.
\end{eqnarray}
Assuming $0 < {\bf q} \ll \lambda_0$, and neglecting ${\bf q}^{\alpha+1}$,
both of the results in Eq.~(\ref{e64}) reduce to $2^{1-\alpha} r C
\lambda_0^{\alpha} {\bf q}$. This analysis shows that the error in
$g_r({\bf q})$ as obtained through the first-order expansion of the
integrand of $g_r({\bf q})$ is of order $\vert {\bf q} \vert^{\alpha+1}$
and {\sl not} of order $\vert {\bf q}\vert^3$. As we have indicated above,
in practice $\alpha$ is small (recall that for the one-dimensional Hubbard
model, $0 < \alpha < 1/8$), so that the range of the ${\bf q}$ values for
which $\vert {\bf q}\vert$ is dominant with respect to $\vert {\bf q}
\vert^{\alpha+1}$ is substantially reduced as compared with the range for
which $\vert {\bf q}\vert$ is dominant with respect to $\vert {\bf q}
\vert^3$. Note in passing that $\vert {\bf q}\vert^{\alpha+1} = \vert
{\bf q}\vert^3$ provided $\alpha=2$ which coincides with the condition
required for the equality of the ${\cal I}_2$'s as evaluated without
relying upon expansion and through (correct) expansion of the integrand
of ${\cal I}_2$.
In $d > 1$, contrary to $d=1$ (in particular when dealing with the
one-dimensional Luttinger model), the radial cut-off parameter $\lambda$
cannot be arbitrarily large. Consequently, the condition $\vert {\bf q}
\vert \ll \lambda_0$ is extremely restrictive, owing to $\lambda_0\ll
\lambda \ll \|{\bf k}_F\|$, and therefore should be relaxed.
Correspondingly, evaluation of integrals subsequent to an expansion of
the pertinent integrands yields results which even in the leading order
deviate from the exact results (see Eqs.~(\ref{e61}), (\ref{e62}) and
(\ref{e64}) above).
Finally, the momentum distribution function in Eq.~(\ref{e60}) is peculiar
in that ${\sf n}_r({\bf k})$ does not allow for a Taylor expansion in a
neighbourhood of ${\bf k}=r k_F$. In the case of the {\sl conventional}
FLs, although ${\sf n}({\bf k})$ is discontinuous at ${\bf k}
={\bf k}_F$, nevertheless it possesses regular (Taylor) expansions to
the left and to the right of ${\bf k}={\bf k}_F$.
\footnote{\label{f19}
Within the framework of the RPA, ${\sf n}({\bf k})$ pertaining to the
uniform system of electrons (in $d=3$) has been shown to possess finite
left and right derivatives at ${\bf k}={\bf k}_F$ (Daniel and Vosko 1960).
On the other hand, for a uniform system of electrons interacting through
a hard-core two-body potential, Belyakov (1961) (see also Sartor and
Mahaux 1980) has obtained an ${\sf n}({\bf k})$ which has logarithmically
divergent derivatives to the left and right of ${\bf k}={\bf k}_F$.
For a discussion of the behaviour of ${\sf n}({\bf k})$ in the close
vicinity of ${\bf k}={\bf k}_F$ see (Farid 1999a). }
Consequently, in cases where one has to do with conventional FLs,
{\sl provided} one takes appropriate care of the discontinuity at ${\bf k}
={\bf k}_F$ of the associated ${\sf n}({\bf k})$, the procedure based
upon expansion of ${\sf n}({\bf k}-{\bf q}/2)-{\sf n}({\bf k}+{\bf q}/2)$
yields correct results to the expected order. One observes that
mathematical procedures that are in general inapplicable, may be applied,
with due care, when dealing with conventional FLs (see Farid 1999a).
The same procedure fails however, in the case of unconventional
FLs or NFLs.
% Sec. V:
\section{Summary and concluding remarks}
In this work we have presented a brief outline of the steps that underlie
a generalised method which has been modelled on a formalism due to DL
(1974) which exactly solves the {\sl one-dimensional} Luttinger (1963) model.
Key elements in both the original DL (1974) method and its generalisation
(MCDC 1998) are the conservation laws and the associated Ward identities.
These identities give rise to some auxiliary relationships which are
{\sl exact} in the one-dimensional Luttinger model, rendering this model
exactly solvable, whereas though {\sl not} exact, these auxiliary
relationships have been asserted to be {\sl asymptotically exact} for
Fermi systems in $d > 1$ in the limit where the scattering of the electrons
in a thin layer circumscribing the Fermi surface is strictly in the
forward-scattering channel (corresponding to scattering processes with
vanishingly-small momentum transfers, that is ${\bf q}\to {\bf 0}$). We
have presented an analysis concerning a crucial relationship between the
irreducible density and current vertex parts, $\Lambda_{\sigma}^0(p;q)$
and ${\bf\Lambda}_{\sigma}(p;q)$ respectively, which is exactly valid in
the case of the one-dimensional Luttinger model and asserted (MCDC 1998)
to be {\sl asymptotically exact} in the forward-scattering limit when
$d > 1$. We have explicitly demonstrated failure of this relationship for
the conventional FL systems in $d=2$. We have further demonstrated that
this relationship is exact, in any arbitrary spatial dimension, {\sl only}
in the trivial case of systems of strictly non-interacting fermions.
\footnote{\label{f20}
Recall that metallic states of non-interacting fermions are {\sl ideal} FLs.}
From the same analysis it follows that for NFLs in $d > 1$ there does
{\sl not} exist even the remotest ground on the basis of which the presumed
asymptotic validity of the mentioned relationship can be justified: here,
the interlocking of energies and momenta in the arguments of the
corresponding self-energies, which by the very fact of being by definition
non-differentiable in neighbourhoods of the Fermi energy and/or the Fermi
momentum (otherwise the associated metallic states would not be NFLs),
stands in the way of obtaining a relationship similar in appearance to
the indicated supposed asymptotic relationship between the irreducible
vertex parts.
We have further explicitly demonstrated that the continuity equation {\sl
only} in $d=1$ gives rise to a direct relationship between the vertex parts
$\Lambda_{\sigma}^{\nu}(p;q)$ and $\Gamma_{\sigma}^{\nu}(p;q)$, $\nu=0,1,
\dots, d$, independent of the analytic properties of the momentum
distribution function ${\sf n}_{\sigma}({\bf k})$. In $d > 1$, continuity
as well as continuous differentiability of this function in the
above-indicated narrow band of the momentum space circumscribing the Fermi
surface {\sl must} be assumed before a similar direct relationship between
the two vertex parts, as in $d=1$, can be deduced. As is well known,
${\sf n}_{\sigma}({\bf k})$ is discontinuous at ${\bf k}={\bf k}_F$ both
for FLs and certain class of NFLs (for the latter see Farid 1999a). For
the case of the one-dimensional Luttinger model, though continuous,
${\sf n}_{\sigma} ({\bf k})$ is {\sl not} differentiable at ${\bf k}
={\bf k}_F$. Therefore, by assuming a similar behaviour for
${\bf n}_{\sigma}({\bf k})$ in $d > 1$ as in $d=1$, we have established
that the indicated simplified relationships between
$\Lambda_{\sigma}^{\nu}(p;q)$ and $\Gamma_{\sigma}^{\nu}(p;q)$, $\nu=0,1,
\dots, d$, do {\sl not} hold in $d > 1$.
Each of the above-mentioned two findings removes ground for validity
in $d > 1$ of the heretofore supposed asymptotic Ward identities in the
forward-scattering limit. This failure concerns both FLs (excluding
strictly non-interacting metallic systems) and NFLs.
In an earlier work (Farid 1999b) we have demonstrated failure of the
Haldane (1992) bosonisation scheme in $d > 1$ both on general grounds
and on grounds of an explicit calculation on a model of interacting
fermions in $d=2$. Within the framework of the Haldane bosonisation scheme,
the explicit expression for the single-particle Green function $G$ takes a
simple form in the space-time representation. It is interesting that
through replacing a screening function that features in this expression
by its random-phase approximation (that is RPA) counterpart, one obtains
an expression which coincides with that obtained within the generalised DL
framework; compare Eq.~(5.65) in (MCDC 1998) with for instance Eq.~(24)
in (Kwon, Houghton and Marston 1995). The use of the RPA within the framework
of the generalised DL method is considered to be justified through reliance
upon the so-called generalised CLT (Kopietz, Hermisson and Sch\"onhammer
1996) which implies that while the RPA yields the {\sl exact} response
function in the case of the one-dimensional Luttinger model (DL 1974)
(see Eq.~(\ref{e56}) above), it yields an {\sl asymptotically exact}
response function in $d > 1$ in the forward-scattering limit. In (Farid
1999b) we have commented on the generalised CLT, however independent of
the true status of this theorem in $d > 1$, and whether one has to do
with FL or NFL systems, from the point of view of our considerations in
the present work the coincidence of the two Green functions is of crucial
importance as it indirectly establishes that indeed the generalised DL (MCDC
1998) formalism is of {\sl no} validity for systems of interacting fermions
in $d > 1$.
To summarise the main conclusions arrived at in this and our earlier
works (Farid 1999a,b), we can state that {\sl none} of the presumed
non-perturbative methods designed thus far for investigating the
existence or otherwise of NFL metallic states in $d > 1$ has proved
capable of dealing with systems of interacting fermion in $d > 1$, not
even in the weakly-interacting limit. Renewed attempts concerning design
and application of non-perturbative methods are therefore called for in
order to settle the long-standing problem with regard to the existence or
otherwise of NFL metallic states in spatial dimensions greater than one.
It is therefore befitting to close this work by mentioning
that Laughlin's (1998) recent statement, that ``... and all attempts
to account for the existence of a Luttinger-liquid at zero temperature
in spatial dimension greater than one have failed.'', must needs be
qualified and in due course perhaps revoked altogether.
%\end{document}
$\hfill\Box$
\vspace{0.33cm}
\section*{Acknowledgements}
\noindent
It is a pleasure for me to thank Professor Peter B. Littlewood and
members of the Theory of Condensed Matter Group for their kind hospitality
at Cavendish Laboratory where this work was completed. I record my
indebtedness to Girton College, Cambridge, for invaluable support. I
extend my thanks to Dr Peter Kopietz for kindly providing me with
references (Kopietz 1997) and (Bohr 1981). I dedicate this
work to the memory of Lady Bertha Jeffreys (1903 - 1999).
%---------------------------------------------
%\end{document}
% Appendix:
\appendix
\section{On dimensional continuation and its application to interacting
Fermi systems}
In this Appendix we first present a brief account of the technique of
dimensional continuation (Wilson 1973, appendix).
\footnote{\label{f21}
For a comprehensive review of this topic see Collins (1984, Chapter 4).}
Subsequently, we point out that the formalism of interacting fermions
even in the limit of strong forward scattering {\sl cannot} be subjected
to the conventional process of dimensional continuation. We consequently
propose an unconventional approach to this process which involves one
continuous free parameter and therefore defines an infinite family of
continuations that correctly interpolate between $d=1$ and $d=2$; they
also correctly interpolate between $d=2$ and $d=3$ for the class of
cylindrically symmetric integrands.
Consider
\begin{equation}
\label{ea1}
{\cal J}_d({\bf q}) {:=} \int_{S} {\rm d}^d{\bf k}\;
f_{\bf q}({\bf k}),
\end{equation}
where $d$ is some real or complex number. Let $\{{\bf e}_1,\dots,
{\bf e}_J\}$ be an {\sl orthonormal} basis spanning the so-called
{\sl parallel space} (to whose complimentary space we refer as {\sl
tangent space}), large enough to accommodate the vector ${\bf q}$
(in general, {\sl all} external momenta $\{{\bf q}_j\}$ upon which
the integrand may depend). We have
\begin{equation}
\label{ea2}
S \equiv S_{\parallel} \cup S_{\perp}.
\end{equation}
With
\begin{equation}
\label{ea3}
{\bf k} = {\bf k}_{\parallel} + {\bf k}_{\perp}
\equiv \sum_{j=1}^J k_{\parallel}^j {\bf e}_j + {\bf k}_{\perp},
\end{equation}
one can therefore write
\begin{equation}
\label{ea4}
{\cal J}_d({\bf q}) = \int_{S_{\parallel}} {\rm d}k_{\parallel}^1\dots
{\rm d}k_{\parallel}^J\; \int_{S_{\perp}} {\rm d}^{d-J} {\bf k}_{\perp}
\; f_{\bf q}({\bf k}).
\end{equation}
By assuming independence of $f_{\bf q}({\bf k})$ upon {\sl direction}
of ${\bf k}_{\perp}$, i.e. by assuming
\begin{equation}
\label{ea5}
f_{\bf q}({\bf k}) \equiv
{\tilde f}_{\bf q}({\bf k}\cdot {\bf k},{\bf k}\cdot {\bf q}),
\end{equation}
one has
\begin{equation}
\label{ea6}
\int_{S_{\perp}} {\rm d}^{d-J} {\bf k}_{\perp}\;
f_{\bf q}({\bf k}) =
K_{d-J}\, \int_{0}^{\infty} {\rm d}\|{\bf k}_{\perp}\|\;
\|{\bf k}_{\perp}\|^{d-J-1}\,
{\tilde f}_{\bf q}({\bf k}\cdot {\bf k},{\bf k}\cdot {\bf q}),
\end{equation}
where
\begin{equation}
\label{ea7}
K_{\nu} {:=} \frac{2 \pi^{\nu/2}}{\Gamma(\nu/2)},
\end{equation}
with $\Gamma(z)$ the gamma function. The value of $K_{\nu}$ in
Eq.~(\ref{ea7}) is equal to the area of the hyper-sphere of unit radius
in $\nu$ dimensions. The $J+1$ integrals over $k_{\parallel}^1,\dots,
k_{\parallel}^J$ and $\|{\bf k}_{\perp}\|$ are conventional and are
evaluated in the usual way. It should be evident that the key element
for reducing the abstract integral in Eq.~(\ref{ea1}) into a conventional
multi-dimensional integral is the assumption specified by the expression
in Eq.~(\ref{ea5}). Below we explicitly consider two cases which clarify
the role played by the dimension of the parallel space, $J$.
\vspace{0.6cm}
\noindent{\it Case I}:
Let $J=1$. Let further $\theta$ denote the planar angle between ${\bf k}$
and ${\bf q}$. With
\begin{equation}
\label{ea8}
\|{\bf k}_{\perp}\| = \|{\bf k}\| \sin\theta,\;\;\;\;
k_{\parallel}^1 = \|{\bf k}\| \cos\theta,\;\;\;
\theta \in [0,\pi],
\end{equation}
one has ${\rm d}k_{\parallel}^1 {\rm d}\|{\bf k}_{\perp}\| = {\rm d}\theta
{\rm d}\|{\bf k}\|\,\|{\bf k}\|$, $\|{\bf k}_{\perp}\|^{d-2} = \|{\bf
k}\|^{d-2} \big(\sin\theta\big)^{d-2}$ and thus (Wilson 1973)
\begin{eqnarray}
\label{ea9}
{\cal J}_d({\bf q}) &=& \int_{-\infty}^{\infty}
{\rm d} k_{\parallel}^1\; K_{d-1}
\int_0^{\infty} {\rm d}\|{\bf k}_{\perp}\|\,
\|{\bf k}_{\perp}\|^{d-2}\,
{\tilde f}_{\bf q}({\bf k}\cdot {\bf k},{\bf k}\cdot {\bf q})\nonumber\\
&\equiv& K_{d-1} \int_{0}^{\pi}
{\rm d}\theta\; \int_0^{\infty} {\rm d}\|{\bf k}\|\;
\|{\bf k}\|^{d-1} \big(\sin\theta\big)^{d-2}\,
{\tilde f}_{\bf q}(\|{\bf k}\|^2,\|{\bf k}\|\,\|{\bf q}\|\cos\theta).
\end{eqnarray}
Since $K_{d-1} \big(\sin\theta\big)^{d-2} \to \delta(\theta) +
\delta(\theta -\pi)$ for $d\downarrow 1$, it follows that in the limit
$d\downarrow 1$ the expression on the RHS of Eq.~(\ref{ea9}) recovers
the pertinent expression for $d=1$, i.e.
\begin{eqnarray}
\label{ea10}
\left. {\cal J}_d({\bf q})\right|_{d\downarrow 1} &=&
\int_0^{\infty} {\rm d}\|{\bf k}\|\,
{\tilde f}_{\bf q}(\|{\bf k}\|^2,\|{\bf k}\|\,\|{\bf q}\|) +
\int_0^{\infty} {\rm d}\|{\bf k}\|\,
{\tilde f}_{\bf q}(\|{\bf k}\|^2,-\|{\bf k}\|\,\|{\bf q}\|)\nonumber\\
&\equiv& \int_{-\infty}^{\infty} {\rm d} k\;
{\tilde f}_{\bf q}(k^2, k\|{\bf q}\|).
\end{eqnarray}
For $d=2$, for which $K_{d-1} = 2$ holds, we have
\begin{eqnarray}
\label{ea11}
\left. {\cal J}_d({\bf q})\right|_{d=2} &=&
2 \int_0^{\pi} {\rm d}\theta\;
\int_0^{\infty} {\rm d}\|{\bf k}\|\; \|{\bf k}\|\,
{\tilde f}_{\bf q}(\|{\bf k}\|^2, \|{\bf k}\|\,\|{\bf q}\|\cos\theta)
\nonumber\\
&\equiv& \int_0^{2\pi} {\rm d}\varphi\;
\int_0^{\infty} {\rm d}\|{\bf k}\|\; \|{\bf k}\|\,
{\tilde f}_{\bf q}(\|{\bf k}\|^2,\|{\bf k}\|\,\|{\bf q}\|\cos\varphi).
\end{eqnarray}
The latter result, whose validity can be readily verified,
\footnote{\label{f22}
To this end employ the variable transformation $\varphi\rightharpoonup
u$ defined through $u = \cos\varphi$, implying ${\rm d}\varphi = \mp
{\rm d} u/(1- u^2)^{1/2}$ for $\varphi\in [0,\pi]$ and $\varphi\in [\pi,
2\pi]$ respectively, and a subsequent reverse transformation
$u\rightharpoonup \varphi$.}
makes explicit that for $d=2$ one indeed recovers the conventional integral
over ${\bf k}$ of ${\tilde f}_{\bf q}(\|{\bf k}\|^2,{\bf k}\cdot {\bf q})$
in terms of the cylindrical-polar coordinates $(\|{\bf k}\|,\varphi)$,
with ${\bf q}$ the polar axis. It should be noted however, that {\sl not}
all two-dimensional integrals can be viewed as the $d\uparrow 2$ limit of
the expression on the RHS of Eq.~(\ref{ea9}).
\vspace{0.6cm}
\noindent{\it Case II}:
Let $J=2$, so that $\{{\bf e}_1,{\bf e}_2\}$ spans the parallel space.
With $\varphi$ denoting the planar angle between ${\bf k}_{\parallel}$
and ${\bf e}_1$ (which without loss of generality we assume to be along
${\bf q}$) and $\theta$ that between ${\bf k}$ and ${\bf k}_{\perp}$
(i.e. ${\bf k}_{\perp}$ is taken as the polar axis), one has
\begin{equation}
\label{ea12}
\|{\bf k}_{\perp}\| = \|{\bf k}\| \cos\theta,\;\;
k_{\parallel}^1 = \|{\bf k}\| \sin\theta\,\cos\varphi,\;\;
k_{\parallel}^2 = \|{\bf k}\| \sin\theta\,\sin\varphi,\;\;\;
\theta\in [0,\pi/2],\;\;\varphi\in [0,2\pi],
\end{equation}
so that ${\rm d}k_{\parallel}^1 {\rm d}k_{\parallel}^2
{\rm d}\|{\bf k}_{\perp}\| = {\rm d}\varphi {\rm d}\theta
{\rm d}\|{\bf k}\|\,\|{\bf k}\|^2\sin\theta$,
$\|{\bf k}_{\perp}\|^{d-3} = \|{\bf k}\|^{d-3}
\big(\cos\theta\big)^{d-3}$ and thus
\begin{eqnarray}
\label{ea13}
{\cal J}_d({\bf q}) &=& \int_{-\infty}^{\infty} {\rm d}k_{\parallel}^1\,
{\rm d}k_{\parallel}^2\; K_{d-2}\, \int_0^{\infty}
{\rm d}\|{\bf k}_{\perp}\|\, \|{\bf k}_{\perp}\|^{d-3}\,
{\tilde f}_{\bf q}({\bf k}\cdot {\bf k}, {\bf k}\cdot {\bf q})\nonumber\\
&=& K_{d-2} \int_0^{2\pi} {\rm d}\varphi\,\int_0^{\pi/2}
{\rm d}\theta\, \big(\cos\theta\big)^{d-3}\,\sin\theta\,\nonumber\\
& &\;\;\;\;\;\;\;\;\;\;\;\times
\int_0^{\infty} {\rm d}\|{\bf k}\|\; \|{\bf k}\|^{d-1}\,
{\tilde f}_{\bf q}(\|{\bf k}\|^2,\|{\bf k}\|\,\|{\bf q}\|
\sin\theta\,\cos\varphi).
\end{eqnarray}
Since $K_{d-2} \big(\cos\theta\big)^{d-3} \to \delta(\theta-
\pi/2)$ for $d\downarrow 2$, it follows that
\begin{equation}
\label{ea14}
\left. {\cal J}_d({\bf q})\right|_{d\downarrow 2}
= \int_0^{2\pi} {\rm d}\varphi\,
\int_0^{\infty} {\rm d}\|{\bf k}\|\, \|{\bf k}\|\,
{\tilde f}_{\bf q}(\|{\bf k}\|^2,\|{\bf k}\|\,\|{\bf q}\|
\cos\varphi),
\end{equation}
which is indeed the expected expression for a two-dimensional integral
in the cylindrical-polar coordinates $(\|{\bf k}\|,\varphi)$. However,
again as in the case considered above (concerning $J=1$), {\sl not} all
two-dimensional integrals over arbitrary functions of ${\bf k}$ can be
obtained as the $d\downarrow 2$ limit of the $d$-dimensional integral
in Eq.~(\ref{ea13}).
\vspace{0.6cm}
The above considerations make explicit that whereas by an appropriate
choice of the dimension of the {\sl parallel space} one can analytically
continue $d$ and obtain the standard integrals for integer dimensions,
nonetheless the applicability of this procedure in an essential way depends
on the requirement that $f_{\bf q}({\bf k})$ in so far as its dependence
upon ${\bf k}$ is concerned, be a function of at most ${\bf k}\cdot {\bf k}$
and ${\bf k}\cdot {\bf q}$. In cases where the dependence of $f_{\bf q}
({\bf k})$ upon ${\bf k}$ is more general, the conventional process of
dimensional continuation may be possible in the limit of small
$\|{\bf q}\|$. In this limit, provided $f_{\bf q}({\bf k})$ be
continuously differentiable in a neighbourhood of ${\bf q}={\bf 0}$, it
may be possible (that is, under the condition that $f_{\bf q}({\bf k})$
to linear order in ${\bf q}$ can be described by a function of ${\bf k}
\cdot {\bf k}$ and ${\bf k}\cdot {\bf q}$) to carry through the process
of dimensional continuation, accurate to linear order in ${\bf q}$.
\footnote{\label{f23}
We note that here we only deal with scalar functions ($f_{\bf q}({\bf k})$)
which are functions of scalar products, however, the technique of
dimensional continuation is also applicable to tensor functions. See
Wilson (1973, appendix) and Collins (1984, Chapter 4).}
This is precisely the case in the generalised DL formalism by MCDC (1998)
(see the last part of \S~II).
\vspace{0.6cm}
It is readily verified that the following expression, whose origin does
{\sl not} lie in the {\sl conventional} formalism for dimensional
continuation (Wilson 1973), is a smooth analytic continuation for $d\in
(1,2]$:
\begin{eqnarray}
\label{ea15}
{\cal J}_d({\bf q}) &{:=}& \frac{K_{d-1}}{\vartheta(d)}\Big\{
\int_0^{\pi} {\rm d}\theta\; \big(\sin\theta\big)^{d-2}\,
\int_0^{\infty} {\rm d}\|{\bf k}\|\; \|{\bf k}\|^{d-1}\,
{\bar f}_{\bf q}(\|{\bf k}\|,\theta)
\nonumber\\ & &\;\;
+\int_{\pi}^{\vartheta(d)\pi} {\rm d}\theta\;
\big(\sin(\theta-\pi)\big)^{d-2}\,
\int_0^{\infty} {\rm d}\|{\bf k}\|\;
\|{\bf k}\|^{d-1}\,
{\bar f}_{\bf q}(\|{\bf k}\|,\theta)\Big\},
\end{eqnarray}
where
\begin{equation}
\label{ea16}
{\bar f}_{\bf q}(\|{\bf k}\|,\theta) \equiv f_{\bf q}({\bf k}),
\end{equation}
with $(\|{\bf k}\|,\theta)$ the cylindrical-polar coordinates of
${\bf k}$, and (see Figure)
\begin{equation}
\label{ea17}
\vartheta(d) {:=} 1 + (3 - d) \exp\big(a \big[1 - 1/(d-1)^{1/a}\big]\big),
\;\;\; \mbox{\rm with}\;\;\ a > 0.
\end{equation}
We note that the integral in Eq.~(\ref{ea15}) together with $\vartheta(d)$
in Eq.~(\ref{ea17}) is also correct for $d=3$ over the space of functions
with cylindrical symmetry; therefore it is also a smooth continuation for
$d\in [2,3]$, however over a restricted set of functions. In this respect
Eq.~(\ref{ea15}) is of limited validity for $d > 2$, in the same way that
Eq.~(\ref{ea9}) is of limited validity for $d > 1$.
The choice for $\vartheta(d)$ in Eq.~(\ref{ea17}) is {\sl not} unique.
However, in choosing any $\vartheta(d)$ interpolating between $\vartheta(1)
=1$ and $\vartheta(2)=2$ (and possibly $\vartheta(3)=1$, etc.) care must
be exercised that in the limit $d\to 1$ the singularities of $K_{d-1}
\big(\sin\theta\big)^{d-2}$ and $K_{d-1}\big(\sin(\theta-\pi)\big)^{d-2}$
for respectively $\theta\uparrow \pi$ and $\theta\downarrow \pi$ do not
yield repeated contributions. This is prevented by the requirement
$\big(\vartheta(d)-1\big)^{d-1} \to 0$ for $d\downarrow 1$; in
consequence of this, for $d\downarrow 1$ the reduction in the width of
the interval $[\pi,\vartheta(d)\pi]$ renders the growth in the integrand,
due to the singularity of $K_{d-1}\big(\sin(\theta-\pi) \big)^{d-2}$ at
$d=1$, ineffective. It can be easily verified that $\vartheta(d)$ in
Eq.~(\ref{ea17}) indeed satisfies this requirement. The freedom in choice
of the continuous positive parameter $a$ in Eq.~(\ref{ea17}) implies that
the dimensional continuation scheme as prescribed by Eqs.~(\ref{ea15}) and
(\ref{ea17}) represents an infinite family of dimensional continuations.
However, in order to recover the expression in Eq.~(\ref{ea9}), applicable
to functions $\{f_{\bf q}({\bf k})\}$ whose dependence upon ${\bf k}$ is
through ${\bf k}\cdot {\bf k}$ and ${\bf k}\cdot {\bf q}$, it is required
that $0< a \ll 1$ (see Figure); in the limit $a\downarrow 0$ the expressions
in Eqs.~(\ref{ea9}) and (\ref{ea15}) coincide.
$\hfill\Box$
%\vfill
%\pagebreak
\begin{references}
\centerline{\sc References}
%\begin{verse}
{\sc Anderson}, P.~W.,
1988, in {\sl Frontiers and Borderlines in Many-Particle Physics},
J.~R. Schrieffer and R.~A. Broglia (editors) (Amsterdam: North Holland),
pp. 1-40;\\
--- 1989, in {\sl Strong Correlation and Superconductivity}, H. Fukuyama,
S. Maekawa and A.~P. Malozemoff (editors) (Berlin: Springer Verlag),
pp. 2-13;\\
--- 1990a, {\sl Phys. Rev. Lett.}, {\bf 64}, 1839;\\
--- 1990b, {\sl Phys. Rev. Lett.}, {\bf 65}, 2306;\\
--- 1991, {\sl Phys. Rev. Lett.}, {\bf 66}, 3226;\\
--- 1993, {\sl Phys. Rev. Lett.}, {\bf 71}, 1220;\\
--- 1997, {\sl THE Theory of Superconductivity in the High-$T_c$
Cuprate Superconductors} (Princeton, NJ: Princeton University Press).\\
{\sc Bares}, P.-A., and {\sc Wen}, X.-G., 1993, {\sl Phys. Rev.} B,
{\bf 48}, 8636.\\
{\sc Belyakov}, V.~A., 1961, {\sl Soviet Phys.} --- {\sl JETP},
{\bf 13}, 850.\\
{\sc Bloch}, F., 1933, {\sl Z. Phys.}, {\bf 81}, 363;\\
--- 1934, {\sl Helv. Phys. Acta}, {\bf 7}, 385.\\
{\sc Bloom}, P., 1975, {\sl Phys. Rev.} B, {\bf 12}, 125.\\
{\sc Bohr}, T., 1981, {\sl The Luttinger Model} Lecture notes in the
series on {\sl Low-Dimensional Statistical Mechanics} (Demark:
{\sc nordita} preprint, {\sc nordita}-81/4).\\
{\sc Castellani}, C., and {\sc Di Castro}, C., 1994, {\sl Physica}
C, {\bf 235}-{\bf 240}, 99.\\
{\sc Castellani}, C., {\sc Di Castro}, C., and {\sc Metzner}, W.,
1994, {\sl Phys. Rev. Lett.}, {\bf 72}, 316.\\
{\sc Coleman}, S., 1975, {\sl Phys. Rev.} B, {\bf 11}, 2088.\\
{\sc Collins}, J.~C., 1984, {\sl Renormalisation} (Cambridge:
Cambridge University Press).\\
{\sc Daniel}, E., and {\sc Vosko}, S.~H., 1960, {\sl Phys. Rev.},
{\bf 120}, 2041.\\
{\sc Di Castro}, C., and {\sc Metzner}, W., 1991, {\sl Phys. Rev.
Lett.}, {\bf 67}, 3852.\\
{\sc DuBois}, D.~F., 1959, {\sl Ann. Phys.} (N.Y.), {\bf 8}, 24.\\
{\sc Dzyaloshinski\v{i}}, I.~E. and {\sc Larkin}, A.~I., 1974,
{\sl Soviet Phys.} --- {\sl JETP}, {\bf 38}, 202.\\
{\sc Engelbrecht}, J.~R., and {\sc Randeria}, M., 1990, {\sl Phys.
Rev. Lett.}, {\bf 65}, 1032;\\
--- 1991, {\sl Phys. Rev. Lett.}, {\bf 66}, 3225.\\
{\sc Everts}, H.~U., and {\sc Schulz}, H., 1974, {\sl Solid State
Commun.}, {\bf 15}, 1413.\\
{\sc Farid}, B., 1999a, {\sl Phil. Mag.} B~{\bf 79} 1097;\\
--- 1999b, submitted.\\
{\sc Fujimoto}, S., 1990, {\sl J. Phys. Soc. Jpn}, {\bf 59}, 2316.\\
{\sc Fukuyama}, H., {\sc Narikiyo}, O., and {\sc Hasegawa}, Y., 1991,
{\sl J. Phys. Soc. Jpn}, {\bf 60}, 372.\\
{\sc Goddard}, P., and {\sc Olive}, D., 1986, {\sl Int. J. Mod. Phys.},
{\bf 1}, 303.\\
{\sc Haldane}, F.~D.~M., 1980, {\sl Phys. Rev. Lett.}, {\bf 45}, 1358;\\
--- 1981, {\sl J. Phys. C}, {\bf 14}, 2585;\\
--- 1992, Proceedings of the International School of Physics
``Enrico Fermi'', Course 121, J.~R. Schrieffer and R.~A. Broglia
(editors) (New York: North Holland), pp. 5 - 30.\\
{\sc Hodges}, C., {\sc Smith}, H., and {\sc Wilkins}, J.~W., 1971,
{\sl Phys. Rev.} B, {\bf 4}, 302.\\
{\sc Houghton}, A., {\sc Kwon}, H.-J., and {\sc Marston}, J.~B.,
1994, {\sl Phys. Rev.} B, {\bf 50}, 1351.\\
{\sc Houghton}, A., {\sc Kwon}, H.-J., {\sc Marston}, J.~B., and
{\sc Shankar}, R., 1994, {\sl J. Phys.} C, {\bf 6}, 4909.\\
{\sc Houghton}, A., and {\sc Marston}, J.~B., 1993, {\sl Phys. Rev.}
B, {\bf 48}, 7790.\\
{\sc Jordan}, P., 1935, {\sl Z. Phys.}, {\bf 93}, 464;\\
--- 1936a, {\sl ibid.}, {\bf 98}, 759;\\
--- 1936b, {\sl ibid.}, {\bf 99}, 109.\\
{\sc Kopietz}, P., 1997, {\sl Bosonization of Interacting Fermions
in Arbitrary Dimensions} in {\sl Lecture Notes in Physics. New Series
m: Monographs} (Berlin: Springer).\\
{\sc Kopietz}, P., {\sc Hermisson}, J., and {\sc Sch\"onhammer}, K.,
1995, {\sl Phys. Rev.} B, {\bf 52}, 10877.\\
{\sc Kr\"onig}, R. de L., 1935a, {\sl Physica}, {\bf 2}, 491;\\
--- 1935b, {\sl ibid.}, 854;\\
--- 1935c, {\sl ibid.}, 968.\\
{\sc Kwon}, H.-J., {\sc Houghton}, A., and {\sc Marston}, J.~B.,
1995, {\sl Phys. Rev.} B, {\bf 52}, 8002.\\
{\sc Laughlin}, R.~B., 1998, {\sl Adv. Phys.}, {\bf 47}, 943.\\
{\sc Littlewood}, P.~B., and {\sc Varma}, C.~M., 1991, {\sl J.
Appl. Phys.}, {\bf 69}, 4979.\\
{\sc Luther}, A., 1979, {\sl Phys. Rev.} B, {\bf 19}, 320.\\
{\sc Luther}, A., and {\sc Peschel}, I., 1974, {\sl Phys. Rev.} B,
{\bf 9}, 2911.\\
{\sc Luttinger}, J.~M., 1961, {\sl Phys. Rev.}, {\bf 121}, 942;\\
--- 1963, {\sl J. Math. Phys.}, {\bf 4}, 1154.\\
{\sc Mattis}, D.~C., and {\sc Lieb}, E.~H., 1965, {\sl J. Math. Phys.},
{\bf 6}, 304.\\
{\sc Metzner}, W., {\sc Castellani}, C., and {\sc Di Castro}, C.,
1998, {\sl Adv. Phys.}, {\bf 47}, 317.\\
{\sc Metzner}, W., and {\sc Di Castro}, C., 1993, {\sl Phys. Rev.} B,
{\bf 47}, 16107.\\
{\sc Sartor}, R., and {\sc Mahaux}, C., 1980, {\sl Phys. Rev.} C,
{\bf 21}, 1546.\\
{\sc Skyrme}, T.~H.~R., 1958, {\sl Proc. Roy. Soc.} A, {\bf 247}, 260;\\
--- 1959, {\sl ibid.}, {\bf252}, 236;\\
--- 1961a, {\sl ibid.}, {\bf 260}, 127;\\
--- 1961b, {\sl ibid.}, {\bf 262}, 237.\\
{\sc Stamp}, P.~C.~E., 1993, {\sl J. Phys.} I Paris, {\bf 3}, 625.\\
{\sc Stone}, M., 1994, {\sl Bosonization} (Singapore: World Scientific).\\
{\sc Tomonaga}, S., 1950, {\sl Prog. Theor. Phys.} (Kyoto), {\bf 5}, 544.\\
{\sc Ueda}, K., and {\sc Rice}, T.~M., 1984, {\sl Phys. Rev.} B,
{\bf 29}, 1514.\\
{\sc Varma}, C.~M., {\sc Littlewood}, P.~B., {\sc Schmitt-Rink}, S.,
{\sc Abrahams}, E., and {\sc Ruckenstein}, A.~E., 1989, {\sl Phys.
Rev. Lett.}, {\bf 63}, 1996;\\
--- 1990 (E), {\sl Phys. Rev. Lett.}, {\bf 64}, 497.\\
{\sc Voit}, J., 1993, {\sl J. Phys.} C, {\bf 5}, 8305;\\
--- 1994, {\sl Rep. Prog. Phys.}, {\bf 57}, 977.\\
{\sc Wilson}, K.~G., 1973, {\sl Phys. Rev.} D, {\bf 7}, 2911.\\
{\sc Wilson}, K.~G., and {\sc Fisher}, M.~E., 1972, {\sl Phys.
Rev. Lett.}, {\bf 28}, 240.\\
%\end{verse}
\end{references}
%\vfill
%\pagebreak
\hrule
\vspace{1.0cm}
% 1.
%\pagebreak
%\clearpage
\begin{figure}[t!]
\protect
\label{fi1}
\centerline{
\psfig{figure=WardId_fig1.eps,width=3.75in} }
\vskip 20pt
\caption{\rm
The function $\vartheta(d)$ as defined in Eq.~(\ref{ea17}) for five distinct
values of the continuous parameter $a$. For $a\downarrow 0$, Eq.~(\ref{ea15})
reduces to the conventional expression in Eq.~(\ref{ea9}) for $d$-dimensional
integration, with $d\in [1,2]$. }
\end{figure}
%
\end{document}